Recent Developments in Automatic

Control Systems
RIVER PUBLISHERS SERIES IN AUTOMATION,
CONTROL AND ROBOTICS
Series Editors:

ISHWAR K. SETHI
Oakland University, USA
TAREK SOBH
University of Bridgeport, USA
FENG QIAO
Shenyang JianZhu University, China

The “River Publishers Series in Automation, Control and Robotics” is a series

of comprehensive academic and professional books which focus on the theory
and applications of automation, control and robotics. The series focuses
on topics ranging from the theory and use of control systems, automation
engineering, robotics and intelligent machines.
Books published in the series include research monographs, edited
volumes, handbooks and textbooks. The books provide professionals,
researchers, educators, and advanced students in the field with an invaluable
insight into the latest research and developments.
Topics covered in the series include, but are by no means restricted to the
following:
• Robots and Intelligent Machines
• Robotics
• Control Systems
• Control Theory
• Automation Engineering

For a list of other books in this series, visit www.riverpublishers.com

 A Comprehensive Study of the
Recent Developments in Automatic
Control Systems

Editors

Yuriy P. Kondratenko
Petro Mohyla Black Sea National University of the Ministry of Education
and Science of Ukraine, Ukraine

Vsevolod M. Kuntsevich
Space Research Institute of the National Academy of Sciences of Ukraine
and State Space Agency of Ukraine, Ukraine

Arkadii A. Chikrii
Glushkov Institute of Cybernetics of the National Academy of Sciences of
Ukraine, Ukraine

Vyacheslav F. Gubarev
Space Research Institute of the National Academy of Sciences of Ukraine
and State Space Agency of Ukraine, Ukraine

River Publishers
 Published 2023 by River Publishers
River Publishers
Alsbjergvej 10, 9260 Gistrup, Denmark
www.riverpublishers.com

Distributed exclusively by Routledge

605 Third Avenue, New York, NY 10017, USA
4 Park Square, Milton Park, Abingdon, Oxon OX14 4RN

Recent Developments in Automatic Control Systems / by Yuriy P. Kon-

dratenko, Vsevolod M. Kuntsevich, Arkadii A. Chikrii.

© 2023 River Publishers. All rights reserved. No part of this publication may
be reproduced, stored in a retrieval systems, or transmitted in any form or by
any means, mechanical, photocopying, recording or otherwise, without prior
written permission of the publishers.

Routledge is an imprint of the Taylor & Francis Group, an informa

business

ISBN 978-87-7022-674-5 (print)

ISBN 978-10-0079-560-8 (online)
ISBN 978-10-0333-922-9 (ebook master)

While every effort is made to provide dependable information, the

publisher, authors, and editors cannot be held responsible for any errors
or omissions.
 Contents

Preface xiii

List of Figures xxi

List of Tables xxix

List of Contributors xxxi

List of Notations and Abbreviations xxxv

Part I Advances in Theoretical Research of Control Systems 1

1 Control of Moving Object Groups in a Conflict Situation 3
A. Chikrii
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Function ω(n, ν), Encirclement by Pshenihnyi, Scheme
of the Method of Resolving Functions . . . . . . . . . . . . 4
1.3 Group Pursuit of a Moving Object . . . . . . . . . . . . . . 9
1.4 Non-fixed Time of Game Termination . . . . . . . . . . . . 14
1.5 The Group Pursuit. Linear State Constraints . . . . . . . . . 17
1.6 Principle of Shortest Broken Line in Successive Pursuit . . . 22
1.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 27
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2 Applications of Variational Analysis to Controlled Sweeping
Processes 31
Boris S. Mordukhovich
2.1 Introduction and Discussions . . . . . . . . . . . . . . . . . 32
2.2 Generalized Differentiation of Variational Analysis . . . . . 36
2.3 Dynamic Optimization via Controlled Moving Sets . . . . . 39
2.4 Sweeping Processes with Controlled Dynamics . . . . . . . 48

v
vi Contents

2.5 Some Applications . . . . . . . . . . . . . . . . . . . . . . 51

2.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . 56
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

3 Robust and Robustly-Adaptive Control of Some Noninvertible

Memoryless Systems 61
L. Zhiteckii and K. Solovchuk
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.2 Problem Statement . . . . . . . . . . . . . . . . . . . . . . 63
3.3 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.4 Robust Nonadaptive Control . . . . . . . . . . . . . . . . . 69
3.5 Robustly-Adaptive Control of Square Systems (Case 1) . . . 76
3.6 Robustly-Adaptive Control of Square Systems (Case 2) . . . 84
3.7 Robustly-Adaptive Control of Nonsquare Systems . . . . . . 86
3.8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 91
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

4 Nonlinear Integral Inequalities and Differential Games

of Avoiding Encounter 97
L.P. Yugay
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 97
4.1.1 On the First Integral Inequalities . . . . . . . . . . . 97
4.1.2 Development of the Theory of Integral Inequalities
(Brief Survey) . . . . . . . . . . . . . . . . . . . . . 98
4.2 Investigation of Nonlinear Hölder Type Integral
Inequality . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
4.2.1 Main Results . . . . . . . . . . . . . . . . . . . . . 100
4.2.2 Applications of Integral Inequalities in Differential
Games of Avoiding an Encounter . . . . . . . . . . 104
4.2.3 Comparison of Results . . . . . . . . . . . . . . . . 105
4.2.4 Remarks . . . . . . . . . . . . . . . . . . . . . . . 106
4.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . 107
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

5 Principle of Time Dilation in Game Problems of Dynamics 113

G.Ts. Chikrii
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 114
 Contents vii

5.2 Statement of the Game Problem. Classic and Modified

Pontryagin’s Condition . . . . . . . . . . . . . . . . . . . . 116
5.3 Method Scheme . . . . . . . . . . . . . . . . . . . . . . . . 118
5.4 Main Statement . . . . . . . . . . . . . . . . . . . . . . . . 120
5.5 Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
5.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 126
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

6 Method of Upper and Lower Resolving Functions for Pursuit

Differential-difference Games with Pure Delay 131
L.V. Baranovska
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 131
6.2 Statement of the Problem . . . . . . . . . . . . . . . . . . . 132
6.3 Scheme of the Method . . . . . . . . . . . . . . . . . . . . 134
6.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 138
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

7 Adaptive Method for the Variational Inequality Problem Over

the Set of Solutions of the Equilibrium Problem 145
Y. Vedel, S. Denisov, and V. Semenov
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 146
7.2 Problem Formulation . . . . . . . . . . . . . . . . . . . . . 147
7.3 Tikhonov–Browder Approximation . . . . . . . . . . . . . . 148
7.4 Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
7.5 Proof of Algorithm Convergence . . . . . . . . . . . . . . . 152
7.6 Algorithm for Two-Level Variational Inequalities . . . . . . 158
7.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Acknowlegdments . . . . . . . . . . . . . . . . . . . . . . . 160
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

Part II Advances in Control Systems Application 165

8 Identification of Complex Systems in the Class of Linear
Regression Models 167
V. Gubarev, S. Melnychuk, and N. Salnikov
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 168
8.2 Active Experiments and Informative Data . . . . . . . . . . 169
8.3 Method of Identification . . . . . . . . . . . . . . . . . . . 173
8.3.1 Model Order Selection . . . . . . . . . . . . . . . . 173
viii Contents

8.3.2 SVD Parametric Identification . . . . . . . . . . . . 176

8.3.3 Total Model Reconstruction . . . . . . . . . . . . . 176
8.3.4 Quasioptimal Model Dimensions . . . . . . . . . . . 177
8.4 Simulation Results . . . . . . . . . . . . . . . . . . . . . . 178
8.4.1 Algorithm of identification . . . . . . . . . . . . . . 179
8.4.2 Excitation of system modes . . . . . . . . . . . . . 181
8.4.3 Effect of Using the Signal Part of Matrices . . . . . 186
8.4.4 Effect of Discarding Data Close to Noise . . . . . . 187
8.4.5 The Best Systems for Identification . . . . . . . . . 188
8.4.6 Model Order Determination . . . . . . . . . . . . . 188
8.4.7 Maximum Dimension of Identifiable Model . . . . . 192
8.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 193
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 194

9 Fuzzy Systems Design: Optimal Selection of Linguistic Terms

Number 197
Oleksiy Kozlov, Yuriy Kondratenko, Oleksandr Skakodub,
and Zbigniew Gomolka
9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 198
9.2 Related Works and Problem Statement . . . . . . . . . . . . 199
9.3 Information Technology for Fuzzy Systems Design
and Structural Optimization with Optimal Selection
of Linguistic Terms Number . . . . . . . . . . . . . . . . . 204
9.4 Design and Structural Optimization of the Fuzzy Control
System for the Quadrotor Unmanned Aerial Vehicle . . . . . 214
9.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 228
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 230

10 Analysis of the Dynamics and Controllability of an Autonomous

Mobile Robot with a Manipulator 241
Ashhepkova Natalja, Zbrutsky Alexander,
and Koshevoy Nicolay
10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 242
10.2 Results of Study of Dynamics and Control of Autonomous
Mobile Robots . . . . . . . . . . . . . . . . . . . . . . . . 244
10.2.1 Problem Statement . . . . . . . . . . . . . . . . . . 244
10.2.2 Design of an Autonomous Mobile Robot
with a Manipulator . . . . . . . . . . . . . . . . . . 246
 Contents ix

10.2.3 Dynamics Analysis . . . . . . . . . . . . . . . . . . 247

10.2.4 Analysis of AMR with M Controllability . . . . . . 253
10.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 256
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . 257
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 257

11 Safe Navigation of an Autonomous Robot in Dynamic

and Unknown Environments 261
Yuriy P. Kondratenko, Arash Roshanineshat,
and Dan Simon
11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 262
11.2 Related Works . . . . . . . . . . . . . . . . . . . . . . . . . 263
11.3 Problem Statement . . . . . . . . . . . . . . . . . . . . . . 267
11.4 Fuzzy based Path Planning . . . . . . . . . . . . . . . . . . 269
11.5 Adjusting Speed . . . . . . . . . . . . . . . . . . . . . . . . 275
11.6 Experiments and Evaluations . . . . . . . . . . . . . . . . . 275
11.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . 278
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 278

12 Algorithmic Procedures Synthesis of Robust-Optimal Control

for Moving Objects 289
V. L. Timchenko, and D.O. Lebedev
12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 290
12.2 Problem Review . . . . . . . . . . . . . . . . . . . . . . . . 294
12.3 Optimal Synthesis . . . . . . . . . . . . . . . . . . . . . . . 296
12.3.1 Optimal Trajectory Planning . . . . . . . . . . . . . 297
12.3.2 Determining Switching Moments of Control
Functions in the Feedback Loops . . . . . . . . . . . 302
12.3.3 Synthesizing Optimal Control Functions
in the Corresponding Feedback Loops . . . . . . . . 302
12.4 Robust Correction . . . . . . . . . . . . . . . . . . . . . . . 305
12.5 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . 309
12.5.1 Sea Vessel Maneuvering . . . . . . . . . . . . . . . 309
12.5.2 Quadrotor UAV Stabilization . . . . . . . . . . . . . 311
12.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 317
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
x Contents

Part III Recent Developments in Collaborative Automation 325

13 Modeling of Cyber–Physical Systems 327
Illya Holovatenko, and Andrii Pysarenko
13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 327
13.2 Review of Modeling Methods for Cyber–Physical Systems . 328
13.3 Problem Formulation . . . . . . . . . . . . . . . . . . . . . 335
13.4 Building Models of a Logistic Cyber–Physical System . . . 336
13.5 Simulation Results . . . . . . . . . . . . . . . . . . . . . . 345
13.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 351
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 352

14 Reliability Control of Technical Systems based on

Canonical Decomposition of Random Sequences 355
I. Atamanyuk, Y. Kondratenko, and M. Solesvik
14.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 356
14.2 Problem Statement . . . . . . . . . . . . . . . . . . . . . . 357
14.3 Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
14.3.1 Forecasting the State of Control Objects . . . . . . . 357
14.3.2 Identification of Random Sequences Model’s
Parameters based on Statistical Goodness-of-fit
Tests . . . . . . . . . . . . . . . . . . . . . . . . . 370
14.3.3 Method of Reliability Control of Technical
Objects . . . . . . . . . . . . . . . . . . . . . . . . 373
14.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 374
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 374

15 Petunin Ellipsoids in Automatic Control Systems Design 379

D.A. Klyushin, S.I. Lyashko, and A.A. Tymoshenko
15.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 380
15.2 Petunin Ellipses and Their Statistical Properties . . . . . . . 381
15.2.1 Petunin Ellipse and Ellipsoids . . . . . . . . . . . . 381
15.2.2 Petunin Ellipses and Linear Constraints . . . . . . . 384
15.2.3 Statistical Properties of Petunin‘s Ellipses . . . . . . 385
15.2.4 Numerical Experiments . . . . . . . . . . . . . . . . 386
15.2.5 Prediction Sets . . . . . . . . . . . . . . . . . . . . 389
15.2.6 Hill’s Assumption A(n) . . . . . . . . . . . . . . . . 389
15.2.7 Testing Statistical Properties for Petunin Ellipses . . 391
15.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 394
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
 Contents xi

16 On Real-Time Calculation of the Rejected Takeoff Distance 399

V.I. Vyshenskyy, A.A. Belousov, and V.V. Kuleshyn
16.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 400
16.2 Dynamic Model of Aircraft Movement along the Runway . . 401
16.3 Analytical Solution of the Aircraft Motion Equation . . . . . 402
16.4 Engine Thrust Modelling During Switching
off and Reverse . . . . . . . . . . . . . . . . . . . . . . . . 402
16.5 Rejected Takeoff Distance Calculation . . . . . . . . . . . . 404
16.6 Interpolation of Aircraft Parameters using Constant
Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
16.7 Estimation of Speed, Path and Time using Fixed
Parameters of the Aircraft . . . . . . . . . . . . . . . . . . . 407
16.8 Computer Simulation . . . . . . . . . . . . . . . . . . . . . 408
16.9 Emulation Results . . . . . . . . . . . . . . . . . . . . . . . 411
16.10Model Verification . . . . . . . . . . . . . . . . . . . . . . 413
16.11Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 413
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 414

17 Automated Control Problem for Dynamic Processes Applied

to Cryptocurrency in Financial Markets 417
Viktor Romanenko, Yurii Miliavskyi, and Heorhii Kantsedal
17.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 418
17.1.1 Problem Statement . . . . . . . . . . . . . . . . . . 419
17.2 Cognitive Mapping of Cryptocurrency Usage in Financial
Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
17.3 Design of Unstable Cryptocurrency Rate Stabilization
System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
17.3.1 Experimental Study of the Stabilization System
of Unstable Cryptocurrency Rate . . . . . . . . . . . 424
17.4 The Problem of Constrained Internal and External
Disturbances Suppression in Control of the Cryptocurrency
CM Impulse Process . . . . . . . . . . . . . . . . . . . . . 427
17.4.1 Algorithm for the State Controller Design
for the CM Impulse Process . . . . . . . . . . . . . 432
17.4.2 Experimental Study of the System of Constrained
Internal and External Disturbances Suppression
in the Cryptocurrency CM Impulse Process
Control . . . . . . . . . . . . . . . . . . . . . . . . 433
xii Contents

17.5 The Problem of Stabilizing the Coordinates

of Cryptocurrency CM Nodes at Given Levels Based
on Varying of Edge Weights and Nodes Coordinates . . . . . 434
17.5.1 Design of a Discrete Controller . . . . . . . . . . . 437
17.5.2 Experimental Studies of the Stabilization
of Cryptocurrency CM Nodes Coordinates
at Given Levels . . . . . . . . . . . . . . . . . . . . 439
17.6 Design of a System for Identifying CM Weighting
Coefficients Based on Recurrent Least Squares Method . . . 440
17.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 442
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 442

Index 445

About the Editors 449

 Preface

This book provides an overview of the recent developments in modern con-

trol systems including new theoretical findings and successful examples of
practical implementation of the control theory in different areas of industrial
and special applications.
The monograph consists of extended versions of the selected papers
which were presented at XXVI International Conference on Automatic
Control “Automation 2020” (October 13-15, 2020, Kyiv, Ukraine) with the
participation of scientists from the different countries (Ukraine, Australia,
United States of America, Poland, Norway, Uzbekistan, Russia and others).
The conference is the main Ukrainian Control Conference, organized by the
Ukrainian Association on Automatic Control (National member organization
of IFAC) and the National Technical University of Ukraine “Igor Sikorsky
Kyiv Polytechnic Institute”.
It is the third monograph in the River Publishers Series in Automation,
Control and Robotics, which is publishing based on the selected papers of the
Ukrainian Control Conferences “Automation”, in particular, first monograph
“Control Systems: Theory and Applications (2018) was published based
on the “Automation – 2017” and second monograph “Advanced Control
Systems: Theory and Applications” – based on the “Automation – 2018”.
The monograph is divided into three main parts: (a) Advances in The-
oretical Research of Control Systems; (b) Advances in Control Systems
Application; (c) Recent Developments in Collaborative Automation.
The first part, “Advances in Theoretical Research on Automatic Control”,
includes seven contributions:
The chapter “Control of Moving Object Groups in Conflict Situation”,
by Chikrii A., presents an overview of the methods for analysis of conflict
situations involving groups of controlled objects on each of the counteracting
sides. The principle of interval decomposition, proposed in the paper, incor-
porates solving typical problems of the target distribution and the problems
of group and successive pursuit. To attack the latter, the method of resolving
functions and Krasovskii’ extremal aiming rule are used. The method of

xiii
xiv Preface

resolving functions makes it feasible to describe the situation of encirclement

of an evader by a group of pursuers, as well as the pursuit under state
constraints.
In “Applications of Variational Analysis to Controlled Sweeping Pro-
cesses”, B. Mordyukhovich discusses recent developments and applications
of variational analysis and generalized differentiation to rather new classes of
optimal control problems governed by discontinuous differential inclusions
that arise in the so-called sweeping processes and the like. Such problems
are particularly important for numerous applications to automatics control
systems, engineering design, hysteresis, etc. To study highly challenging
control problems of this type, author develops appropriate constructions
of the method of discrete approximations, which allow to investigate vari-
ous theoretical and numerical aspects of optimal control for discontinuous
differential systems.
L. Zhiteckii and K. Solovchuk in the chapter “Robust and Robustly-
Adaptive Control of Some Noninvertible Memoryless Systems” discuss the
robust control of some uncertain multivariable memoryless (static) discrete-
time systems in the presence of arbitrary unmeasurable bounded disturbances,
whose bounds may be unknown, in general. The non-adaptive robust con-
trol approach applicable to the case when the lengths of these intervals
are relatively short is proposed. The robustly-adaptive control utilizing a
modification of inverse-model concept is designed to cope with large para-
metric uncertainty. The asymptotic properties of both non-adaptive and
adaptive feedback control systems are derived. To support the theoretic study,
numerical examples and simulation results are given.
The chapter “Nonlinear Integral Inequalities and Differential Games of
Avoiding Encounter “, by L.P. Yugay, presents a survey of the application of
integral inequalities in various theoretical branches of mathematics and their
applied fields. New integral inequalities of the HÃűlder type are introduced
and corresponding estimates are proved. Also the possibilities of applications
of proved results in nonlinear differential games of avoiding encounter are
indicated.
In “Principle of time stretching for motion control in condition of con-
flict”, G.Ts. Chikrii discusses linear differential game of approaching a
cylindrical terminal set. The research is doing in the frames of the Method
of Resolving Functions. The gist of the method consists in constructing
some scalar functions associated with the parameters of a conflict-controlled
process and characterizing the gain of the first player at each moment of
time. If the total gain achieves the predetermined value then this means
 Preface xv

that the game terminates in a certain guaranteed time. In so doing, the

pursuer’s control, realizing the game goal, is constructed on the basis he of
the Filippov-Castaing theorem on measurable choice.
L.V. Baranovska in the chapter “Method of Upper and Lower Resolv-
ing Functions for Pursuit Differential-Difference Games with Pure Delay”
considers a modification of the method of resolving functions for differential-
difference games with pure delay is considered. The lower and upper
resolving functions. For the local game of approach, the scheme of a method
is constructed, sufficient conditions for the completion of the game are
formulated.
The chapter “Adaptive method for variational inequality over the set of
solutions of equilibrium problem”, by Vedel Y.I., Denisov S.V. and Semenov
V.V., considers a two-level problem: a variational inequality over the set of
solutions of the equilibrium problem. An example of such a problem is the
search for the normal Nash equilibrium. To solve this problem, authors pro-
pose an iterative algorithm that combines the ideas of a two-stage proximal
method, adaptability, and iterative regularization. In contrast to the previously
applied rules for choosing the step size, the proposed algorithm does not
calculate the values of the bifunction at additional points; it does not require
knowledge of information about the Lipschitz constants of the bifunction,
the Lipschitz constant and strong monotonicity constant of the operator. For
monotone bifunctions of Lipschitz type and strongly monotone Lipschitz
continuous operators, a theorem on the strong convergence of the algorithm
is proved. It is shown that the proposed algorithm is applicable to monotone
two-level variational inequalities in Hilbert spaces.

The second part, “Advances in Control Systems Applications”, includes five

contributions:
Gubarev V., Salnikov N. and Melnychuk S. in the chapter “Identification
of complex systems in the class of linear regression models” discuss a general
approach to identification within the regression model class, which searches
for approximate solutions that are consistent in accuracy with errors in the
data. Since identification problems in complex systems are often incorrectly
posed, the proposed method includes regularization that provides practically
suitable solutions.
In ”Fuzzy Systems Design: Optimal Selection of Linguistic Terms Num-
ber”, Kozlov O., Kondratenko Y., Skakodub O. and Gomolka Z. focus on
the development and research of the advanced information technology for
fuzzy systems design and structural optimization with optimal selection
xvi Preface

of linguistic terms number. The application of the proposed information

technology allows increasing the fuzzy system’s performance and accuracy,
reducing the computational costs spent on the rule base composing and
parameters optimization, as well as simplifying its further hardware and soft-
ware implementation. The effectiveness study of the developed information
technology is carried out on a specific example, in particular, when designing
and optimizing the fuzzy control system for the quadrotor unmanned aerial
vehicle. The obtained results of the conducted research confirm the high
effectiveness of the proposed information technology and the feasibility
of its use at creating various types of fuzzy control and decision-making
systems.
The chapter “Analysis of the Dynamics and Controllability of
Autonomous Mobile Robot with a Manipulator”, by Ashchepkova N., Zbrut-
sky A. and Koshevoy N., presents the results of the study of the dynamics
and control of an autonomous mobile robot with a manipulator. An all-wheel
drive model with a four-wheeled chassis layout and equipped with a four-
link manipulator is considered. The design of the manipulator consists of a
docking ring rotating around a vertical axis and three rod links of the arm,
connected by rotary kinematic pairs of the fifth class. The mathematical
model of controlled movement of autonomous mobile robot with a manip-
ulator is proposed. With relative motion of the manipulator, the tensor of
inertia of an autonomous mobile robot in the coordinate system associated
with the chassis is non-diagonality and non-stationarity. Simplifications of
the mathematical model for several driving modes are substantiated. An
algorithm for integrating the equations of dynamics is developed, taking
into account certain features. Based on the results of mathematical model-
ing, an algorithm for adaptive control of autonomous mobile robot with a
manipulator is proposed.
Kondratenko Y., Roshanineshat A. and Simon D. in chapter “Safe Naviga-
tion of an Autonomous Robot in Dynamic and Unknown Environments” focus
on the problem of fuzzy control system design for mobile robots operating in
dynamic environments with obstacles. The decision-making method, based
on the transformation of radar-sensor information to fuzzy set “Obstacles”
and robot’s heading course to fuzzy set “Direction to Target Point” is consid-
ered. Different scenarios of working environment (with static and dynamic
obstacles) and adjusted speed of the autonomous robot as well as peculiarities
of “robot-obstacle” interaction are discussed in detail. The simulation results
confirm the efficiency and safety of the robot’s autonomous navigation in
dynamic, a priori unknown environments.
 Preface xvii

The chapter “Algorithmic Procedures Synthesis of Robust-Optimal Con-

trol for Moving Objects”, by Timchenko V.L. and Lebedev D.O., considers
the control of transient processes of a wide class of moving objects described
by ordinary nonlinear differential equations with robust-optimal systems and
variable structure. The solution of this problem is suggested using gen-
eral algorithmic procedure for constructing optimal trajectories, determining
switching moments and synthesizing control functions for multidimensional
systems. The control of the mismatch between the trajectory of a physical
object and the optimal model calculated trajectory allows to take into account
the values of the optimal control and to form a robust subsystem, which
provides invariance to incomplete information about the moving object. The
simulation of a quadrotor UAV stabilization and a sea vessel maneuvering
under conditions of external disturbances and parametric noise demonstrates
the required control accuracy in the vicinity of the formed optimal trajectory
of controlled coordinates. The creation of algorithms and circuit solutions
for the synthesis of robust-optimal variable-structure systems for multidi-
mensional nonlinear moving objects has practical significance as a basis
for formation of automated procedures for the synthesis of robust-optimal
systems and development of software tools for various types of moving object
control systems.
The third part, “Recent Developments in Collaborative Automation”, con-
sists of five contributions:
In “Modeling of Cyber–Physical Systems”, Holovatenko I. and Pysarenko
A. present the current research in the field of constructing cyber-physical
systems. For validation of the designed system authors came up with a live
scenario from the logistical systems: the autonomous object delivers goods
from one warehouse to another taking into account obstacles along the way.
The system should proactively react and recalculate the optimal path. This
will lead to the timely delivery of goods to the destination.
Atamanyuk I., Kondratenko Y. and Solesvik M. in the chapter “Reliability
Control of the Technical Systems based on Canonical Decomposition of
Random Sequences” consider a method for determining the suitability of a
technical object for further operation based on the information about the cur-
rent state and the history of its functioning during an arbitrary period of time.
The check is carried out on the basis of the analysis of the state of the object
under study at future points in time and the value of the posterior probability
of no-failure operation. The method for predicting the individual reliability
of objects is based on the canonical decomposition of a random sequence
xviii Preface

describing the change in the value of the controlled parameter over time. As
an estimate of the future state, the conditional mathematical expectation is
used; to estimate the probability of no-failure operation, statistical modeling
of a random sequence beyond the observation area is performed. The method
can also be used to control the reliability of the objects that are characterized
by many parameters.
The chapter “Petunin Ellipsoids in Automatic Control Systems Design”,
by Klyushin D.A., Lyashko S.I. and Tymoshenko A.A., proposes a new way
of describing the optimization domain with linear constraints in the design
space using Petunin ellipsoids. The statistical properties of these ellipsoids
are given, making them an effective tool for describing data in the design
space. Authors consider a new distribution-free approach to computing ellip-
soidal conformal prediction set based on Hill’s assumption and the Petunin
ellipsoids. The resulting prediction set is a Petunin ellipsoid with a precise
confidence level depending on the number of points only.
In ”On Real-Time Calculation of the Rejected Takeoff Distance”, Vyshen-
skyy V.I., Belousov A.A. and Kuleshyn V.V. focus on a rejected takeoff in
the situation when it is decided to cancel the takeoff of an airplane. The
goal of the authors’ investigation is to obtain real-time estimations of the
distance to a full stop. The proposed real-time algorithm for the calculation
of the distance of the interrupted flight uses the step-wise interpolation of the
thrust. Using this interpolation, the Riccati equation of the aircraft motion
has explicit solutions. Besides using known time intervals of interpolation,
the time to reach a given speed can be explicitly calculated. To check and
study the proposed algorithm for estimating the rejected takeoff distance in
real time, a computer simulation system was created.
Romanenko V., Miliavskyi Y. and Kantsedal G. in the chapter “Automated
Control Problem for Dynamic Processes Applied to Cryptocurrency in Finan-
cial Markets” develop a cognitive map (CM) of the use of cryptocurrency
in the financial market and describe the dynamic model of CM impulse pro-
cesses as a difference equation system (Roberts equations). The authors chose
an external control vector for the CM impulse process provided by means of
CM nodes varying. A closed-loop CM impulse process control system was
implemented. This system includes a multidimensional discrete controller,
based on automated control theory methods, that generates a selected control
vector and directly affects respective CM nodes by varying their coordi-
nates. The authors solved three problems of a discrete controller design for
automated dynamic process control applied to cryptocurrency in financial
markets. The first problem is unstable cryptocurrency rate stabilization based
 Preface xix

on modal CM impulse process control. The second problem is constrained

external and internal disturbances suppression during CM impulse processes
control based on the invariant ellipsoid method. The third problem is min-
imizing the generalized variance of CM nodes coordinates and controls for
stabilizing coordinates at given levels.
The chapters of the monograph have been structured to provide an easy-
to-follow introduction to the topics that are addressed, including the most
relevant references, so that anyone interested in this field can get started in
the area.
This book may be useful for researchers and students who are inter-
esting in recent developments of the modern control systems, robust adap-
tive systems, optimal control, fuzzy control, motion control, identification,
modelling, differential games, evolutionary optimization, reliability control,
security control, intelligent robotics and cyber–physical systems.
We would like to express our deep appreciation to all authors for their
contributions as well as to reviewers for their timely and interesting comments
and suggestions. We certainly look forward to working with all contributors
again in nearby future.
July 31, 2021
Editors:

Yuriy P. Kondratenko, Prof., Dr.Sc., Corresponding Academician of the

Royal European Academy of Doctors – Barcelona 1914 (Spain), Petro
Mohyla Black Sea National University, Ukraine
Vsevolod M. Kuntsevich, Prof., Dr.Sc., Academician of the National
Academy of Sciences of Ukraine, Space Research Institute of the National
Academy of Sciences and State Space Agency, Ukraine
Arkadii A. Chikrii, Prof., Dr.Sc., Academician of the National Academy of
Sciences of Ukraine, Glushkov Institute of Cybernetics of the National
Academy of Sciences, Ukraine
Vyacheslav F. Gubarev, Prof., Dr.Sc., Corresponding Member of the
National Academy of Sciences of Ukraine, Space Research Institute of the
National Academy of Sciences and State Space Agency, Ukraine
 List of Figures

Figure 3.1 Configuration of the pseudoinverse model-based

feedback control system without uncertainty . . . . 65
Figure 3.2 Configuration of the pseudoinverse model-based
feedback control system with parametric
uncertainty . . . . . . . . . . . . . . . . . . . . . 66
Figure 3.3 Set Eu of equilibrium points in Example 3.1 . . . . 68
Figure 3.4 Geometric interpretation of the condition . . . . . . 69
Figure 3.5 Geometric interpretation of the condition (3.29) . . 73
Figure 3.6 Choice of matrix B0 satisfying the condition
rank B0 = 1 (possible cases): (a) there are B0 s
belonging to Ξ; (b) there are no B0 s belonging to
Ξ (3.29) . . . . . . . . . . . . . . . . . . . . . . . 74
Figure 3.7 Equilibrium states related to the conditions of
Example 3.3 . . . . . . . . . . . . . . . . . . . . . 76
Figure 3.8 Robust pseudoinverse model-based controller:
(a) Euclidean norm ||un ||2 of control input vectors;
(b) Euclidean norms ||yn ||2 , ||y 0 ||2 of output vectors
(solid line) and desired output vectors (dashed line),
respectively . . . . . . . . . . . . . . . . . . . . . 77
Figure 3.9 The Gerŝgorin discs for r=2 in the case
(1)
|β (2) | < |β | . . . . . . . . . . . . . . . . . . . . 78
Figure 3.10 The Gerŝgorin discs for r=2 in the case
(2)
|β | < |β (1) | . . . . . . . . . . . . . . . . . . . . 78
Figure 3.11 Configuration of adaptive stabilization system . . . 82
Figure 3.12 Estimated parameters of the fictitious plant (solid
line) and the true plant (dashed line) . . . . . . . . 83
Figure 3.13 Performance of the robustly-adaptive controller in
Simulation experiment 3.2: (a) Euclidean norm of
input control vector; (b) Euclidean norms of output
vector . . . . . . . . . . . . . . . . . . . . . . . . 83

xxi
xxii List of Figures

(11) (12) (21) (22)

Figure 3.14 Estimates b̃n , b̃n , b̃n and b̃n of Simulation
experiment 18.3 . . . . . . . . . . . . . . . . . . . 85
(1) (2)
Figure 3.15 Estimates εn /2 and εn /2 of Simulation
experiment 3.3 . . . . . . . . . . . . . . . . . . . 85
(1) (2)
Figure 3.16 Control inputs un and un of Simulation
experiment 3.3 . . . . . . . . . . . . . . . . . . . 86
(1) (2)
Figure 3.17 Outputs yn and yn of Simulation experiment 3.3 86
Figure 3.18 Variables describing the adaptive estimation
processes . . . . . . . . . . . . . . . . . . . . . . 90
Figure 3.19 The behaviour of the control system: (a) the
current number k of control un [k] chosen from
{un [1], un [2], un [3]} at given n; (b) the 1-norm
of control vector; (c) the 1-norm of output vector
in adaptive case (solid line) and in nonadaptive
optimal case (dashed line) . . . . . . . . . . . . . 90
Figure 8.1 State of a complex mode under harmonic excitation
with different initial phases . . . . . . . . . . . . . 182
Figure 8.2 State of a complex mode under harmonic excitation
of different frequencies . . . . . . . . . . . . . . . 183
Figure 8.3 State of a complex mode during relaxation after
harmonic excitation . . . . . . . . . . . . . . . . . 183
Figure 8.4 State of a complex mode during constant excitation
and relaxation . . . . . . . . . . . . . . . . . . . . 184
Figure 8.5 The norm of the state vector of modes under
harmonic excitation . . . . . . . . . . . . . . . . . 184
Figure 8.6 The norm of the state vector of modes under
constant excitation . . . . . . . . . . . . . . . . . 185
Figure 8.7 The norm of the state vector of modes
under rectangular pulse and on relaxation . . . . . 185
Figure 8.8 The maximum of norm of the modes state vectors
under harmonical input . . . . . . . . . . . . . . . 186
Figure 8.9 Average number of rows in overdetermined systems
for a and b (left), the quality of the resulting model
for different implementations of random
noise (right) . . . . . . . . . . . . . . . . . . . . . 188
Figure 8.10 First singular values of the matrix Yrelax for systems
of different order . . . . . . . . . . . . . . . . . . 190
 List of Figures xxiii

Figure 8.11 First singular values of the matrix Ỹrelax for systems
of different order. Case of relative error bound
εY = 10−3 . . . . . . . . . . . . . . . . . . . . . 190
Figure 8.12 First normalized singular values of the matrix Ỹrelax
for systems of different order. Case of relative error
bound εY = 10−3 . . . . . . . . . . . . . . . . . . 191
Figure 8.13 Relative max norm of the matrix YrelaxN . . . . . . 191
Figure 8.14 Accuracy of the model, depending on the dimension
and the noise . . . . . . . . . . . . . . . . . . . . 193
Figure 9.1 Structure of the generalized MIMO fuzzy control/
decision-making system with structural optimiza-
tion unit . . . . . . . . . . . . . . . . . . . . . . . 202
Figure 9.2 Flowchart of the first modification of the informa-
tion technology for FS structural optimization based
on optimal selection of linguistic terms number . . 210
Figure 9.3 Flowchart of the second modification of the infor-
mation technology for FS structural optimization
based on optimal selection of linguistic terms
number . . . . . . . . . . . . . . . . . . . . . . . 212
Figure 9.4 Flowchart of the stage of design of the RB and
calculation of the complex objective function for the
FS with the current vector S . . . . . . . . . . . . 213
Figure 9.5 The fuzzy control system of the QUAV flight
altitude . . . . . . . . . . . . . . . . . . . . . . . 217
Figure 9.6 Complex relief of the mountainous terrain presented
by the dependence zT (x) . . . . . . . . . . . . . . 217
Figure 9.7 Changing curves of the complex objective func-
tion J C and the criteria J 1 , J 2 values during the
structural optimization process (first modification) . 220
Figure 9.8 Changing curves of the complex objective func-
tion J C and the criteria J 1 , J 2 values during the
structural optimization process (second modification) 223
Figure 9.9 QUAV’s flight trajectories over the mountainous
terrain . . . . . . . . . . . . . . . . . . . . . . . . 225
Figure 9.10 Detailed QUAV’s flight trajectories (z = –2. . . 12 m;
x = 0. . . 100 m) . . . . . . . . . . . . . . . . . . . 226
Figure 9.11 Detailed QUAV’s flight trajectories (z = 5. . . 35 m; x
= 780. . . 900 m) . . . . . . . . . . . . . . . . . . . 226
xxiv List of Figures

Figure 9.12 Transients of the altitude control system

of the QUAV with developed FCs . . . . . . . . . 227
Figure 9.13 Detailed transients of the altitude control system
(t = 0. . . 40 s) . . . . . . . . . . . . . . . . . . . . 227
Figure 9.14 Detailed transients of the altitude control system
(zR = 3.5. . . 6.5 m; t = 100. . . 220 s) . . . . . . . . . 228
Figure 10.1 Diagram of the AMR design with
a manipulator . . . . . . . . . . . . . . . . . . . . 246
Figure 10.2 Relay static characteristic . . . . . . . . . . . . . . 251
Figure 10.3 Algorithm for solving the problem of dynamics of
AMR with M by the Runge–Kutta–Fehlberg
method . . . . . . . . . . . . . . . . . . . . . . . 252
Figure 10.4 Block diagram of the AMR adaptive control
algorithm with M . . . . . . . . . . . . . . . . . . 255
Figure 11.1 The scenario of mobile robot’s working
environment with obstacles . . . . . . . . . . . . . 268
Figure 11.2 The radar-beam-diagram of the robot sensor system
in point A on the motion trajectory, presented in
Figure 11.1 . . . . . . . . . . . . . . . . . . . . . 269
Figure 11.3 The fuzzy set ∼ E, which illustrates the proposed
fuzzy interpretation of the radar-beam-diagram,
presented in Figure 11.2 . . . . . . . . . . . . . . 271
Figure 11.4 The fuzzy set G ∼
with rectangular membership
function (11.2) . . . . . . . . . . . . . . . . . . . 272
Figure 11.5 The resulting fuzzy set R ∼
with one-interval core for
the determination of mobile robot’s heading course
according to (11.3), (11.4) . . . . . . . . . . . . . 273
Figure 11.6 The multimodal resulting fuzzy set R ∼
with
multi-intervals core . . . . . . . . . . . . . . . . . 273
Figure 11.7 The highlighted area for defuzzification . . . . . . 275
Figure 11.8 The robot radar division for different speeds . . . . 276
Figure 11.9 Maps of unknown environments used
for experiments . . . . . . . . . . . . . . . . . . . 277
Figure 12.1 Algorithmic procedure A for constructing
the optimal trajectory in terms of maximum
speed . . . . . . . . . . . . . . . . . . . . . . . . 303
Figure 12.2 Algorithmic procedure B for the problem
of minimum energy consumption . . . . . . . . . . 304
 List of Figures xxv

Figure 12.3 Block diagram of the implementation of control in

the form of time polynomials (F(t) is not
considered; the index X for the matrices A,
B is omitted) . . . . . . . . . . . . . . . . . . . . 304
Figure 12.4 Block diagram of the implementation of control in
the form of integrating functions in the feedback . . 306
Figure 12.5 The structure of a linear and stationary
robust-optimal system . . . . . . . . . . . . . . . . 307
Figure 12.6 The structural implementation of a robust-optimal
control system of variable structure . . . . . . . . . 309
Figure 12.7 Course angle over time . . . . . . . . . . . . . . . 310
Figure 12.8 Rudder angle over time . . . . . . . . . . . . . . . 311
Figure 12.9 Course angle error over time . . . . . . . . . . . . 311
Figure 12.10 Simulink-based robust-optimal stabilization scheme
for a quadrotor UAV on a given trajectory . . . . . 313
Figure 12.11 Errors of the higher coordinates over time . . . . . 314
Figure 12.12 Velocity of the higher coordinates over time . . . . 314
Figure 12.13 Lower coordinates of angular orientation over
time . . . . . . . . . . . . . . . . . . . . . . . . . 314
Figure 12.14 Comparison of control accuracy between optimal
control with variable structure and PID control over
time . . . . . . . . . . . . . . . . . . . . . . . . . 315
Figure 12.15 Comparison of energy consumption between
optimal control with variable structure and PID
control over time . . . . . . . . . . . . . . . . . . 315
Figure 12.16 “Error tubes” of the main coordinates x, y, z;
Vw = 10 m/s . . . . . . . . . . . . . . . . . . . . . 316
Figure 12.17 “Error tubes” of the main coordinates x, y, z;
Vw = 3 m/s . . . . . . . . . . . . . . . . . . . . . 316
Figure 13.2 Example of a composite cyber–physical system . . 329
Figure 13.3 Metamodel in UML notation . . . . . . . . . . . . 330
Figure 13.4 A closed-loop of a cyber–physical system . . . . . 332
Figure 13.5 An example of inter-component interaction within
a cyber–physical system . . . . . . . . . . . . . . 336
Figure 13.6 The metamodel of a cyber–physical system . . . . 337
Figure 13.7 BPMN4CPS Cyber–Physical System Notation . . . 338
Figure 13.8 Primary trajectory obtained by ODG at the initial
moment of time . . . . . . . . . . . . . . . . . . . 345
Figure 13.9 Script for constructing the primary trajectory . . . 346
xxvi List of Figures

Figure 13.10 The state of the cyber–physical system at the

moment the radar detects an obstacle (the state of
the radar (a), the state of the ODG (b) . . . . . . . 347
Figure 13.11 MATLAB script responsible for the operation of the
radar and movement of the ODG . . . . . . . . . . 348
Figure 13.12 New logistics route recalculated due
to the occurrence of an obstacle . . . . . . . . . . . 348
Figure 13.13 MATLAB script showing the characteristics of the
radar and the obstacle tracking algorithm . . . . . . 349
Figure 13.14 The state of the cyber–physical system
after receiving a new logistic route (the state of the
radar (a), the state of the ODG (b)) . . . . . . . . . 350
Figure 14.1 Block-diagram of the calculation algorithm for the
parameters Dλ (ν), ρ(λ, ν; h, i) of the polynomial
power of canonical expansion (14.3) . . . . . . . . 360
Figure 14.2 Generative procedure for future values of a random
sequence based on the algorithm (14.26) . . . . . . 368
Figure 15.1 Petunin rectangle Π . . . . . . . . . . . . . . . . . 382
Figure 15.2 Petunin circle C . . . . . . . . . . . . . . . . . . . 383
Figure 15.3 Petunin ellipse E . . . . . . . . . . . . . . . . . . 383
Figure 15.4 Uniformly distributed points with square
constraints . . . . . . . . . . . . . . . . . . . . . . 387
Figure 15.5 Gaussian distributed points with square
constraints . . . . . . . . . . . . . . . . . . . . . . 387
Figure 15.6 Uniformly distributed point with pentagonal
constraints . . . . . . . . . . . . . . . . . . . . . . 387
Figure 15.7 Uniformly distributed point with hexagonal
constraints . . . . . . . . . . . . . . . . . . . . . . 388
Figure 15.8 Area of Petunin ellipses covering 100 uniformly
and Gaussian distributed points after dropping
one point per step . . . . . . . . . . . . . . . . . . 388
Figure 15.9 Ellipse covering point with uniform (left)
and Gaussian (right) distribution . . . . . . . . . . 392
Figure 15.10 Areas of the ellipses covering point with uniform
(left) and Gaussian (right) distribution . . . . . . . 392
Figure 16.1 Dependence of engine thrust on time . . . . . . . . 404
Figure 16.2 User interface . . . . . . . . . . . . . . . . . . . . 410
Figure 16.3 Emulation and integration results . . . . . . . . . . 411
Figure 16.4 Rejected takeoff distance, velocity, and thrust . . . 412
 List of Figures xxvii

Figure 17.1 CM cryptocurrency usage in financial markets . . . 421

Figure 17.2 Results of modelling for the CM . . . . . . . . . . 427
Figure 17.3 CM coordinates and controls changes . . . . . . . 436
Figure 17.4 Results of applying the control actions . . . . . . . 440
 List of Tables

Table 8.1 Maximum and average errors of models

for εU = 10−4 and εY = 10−4 . . . . . . . . . . . . 187
Table 8.2 Maximum and average errors of models
for εU = 10−6 and εY = 10−4 . . . . . . . . . . . . 187
Table 8.3 The worst and average quality of models . . . . . . . 189
Table 8.4 Quality of models found . . . . . . . . . . . . . . . . 192
Table 9.1 Optimization results using the first modification . . . 219
Table 9.2 RB fragment for the vector Sopt obtained using
the first modification . . . . . . . . . . . . . . . . . . 221
Table 9.3 Optimization results using the second
modification . . . . . . . . . . . . . . . . . . . . . . 222
Table 9.4 RB fragment for the vector Sbest obtained by the
second modification . . . . . . . . . . . . . . . . . . 224
Table 10.1 Results of mathematical modeling of AMR dynamics
without taking into account the non-diagonality
and non-stationarity of the tensor of inertia . . . . . . 253
Table 10.2 Results of mathematical modeling of AMR dynamics
taking into account the non-diagonality
and non-stationarity of the tensor of inertia . . . . . . 253
Table 11.1 Table of Notations . . . . . . . . . . . . . . . . . . . 268
Table 11.2 Table of the robot speed . . . . . . . . . . . . . . . . 276
Table 11.3 Result of experiments . . . . . . . . . . . . . . . . . 276
Table 13.1 Physical definitions of ODG . . . . . . . . . . . . . . 340
Table 13.2 Optimal tracking system notation . . . . . . . . . . . 341
Table 15.1 The number of uniformly distributed points dropped
out and the significance level of the Petunin ellipse
satisfying square constraints . . . . . . . . . . . . . . 386
Table 15.2 The number of Gaussian distributed points dropped
out and the significance level of the Petunin ellipse
satisfying square constraints . . . . . . . . . . . . . . 386

xxix
xxx List of Tables

Table 15.3 The results of testing confidence level of the Petunin

ellipse for uniform distribution . . . . . . . . . . . . 393
Table 16.1 Comparison of emulation and integration . . . . . . . 412
Table 16.2 Comparison of emulation and nomograms . . . . . . 413
 List of Contributors

Ashhepkova, Natalja, Department of Mechanotronics, Oles Honchar

Dnipro National University, Ave. Gagarin, 72, Dnipro, Ukraine, 49010;
E-mail: [email protected], ORCID: http:// orcid.org/ 0000
-0002-1870-1062
Atamanyuk, I., Warsaw University of Life Sciences, Nowoursynowska 166,
02-787 Warsaw, Poland and Mykolayiv National Agrarian
University, Georgy Gongadze street, Mykolaiv, 9, 54020, Ukraine;
E-mail:[email protected]
Baranovska, Lesia, Institute of Applied System Analysis of the National
Technical University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”,
37 Peremohy av., Kyiv, 03056, Ukraine; E-mail: [email protected]
Belousov, A.A., V.M. Glushkov Institute of Cybernetics of National Academy
of Sciences of Ukraine, Kyiv, Ukraine; E-mail: [email protected]
Chikrii, A., Glushkov Institute of Cybernetics of the National Academy of
Sciences of Ukraine, 40 Glushkov Avenue, 03187, Kyiv, Ukraine;
E-mail: [email protected]
Chikrii, G.Ts., Glushkov Institute of Cybernetics of the National Academy of
Sciences of Ukraine, 40 Glushkov Avenue, 03187, Kyiv-187, Ukraine;
E-mail: [email protected]
Denisov, S., Faculty of Computer Science and Cybernetics, Taras Shevchenko
National University of Kyiv, Kyiv, Ukraine; E-mail: [email protected]
Gomolka, Zbigniew, University of Rzeszow, Department of Computer Engi-
neering, Rzeszow, Poland; E-mail: [email protected]
Gubarev, V., Space Research Institute, National Academy of Sciences of
Ukraine and State Space Agency of Ukraine, 40, b. 4/1 Glushkov av., Kyiv,
03187, Ukraine

xxxi
xxxii List of Contributors

Holovatenko, Illya, Department of Information Systems and Technologies,

National Technical University of Ukraine “Igor Sikorsky Kyiv Polytechnic
Institute”, Kyiv, Ukraine; E-mail: [email protected]
Kantsedal, Heorhii, Institute of Applied System Analysis of National Techni-
cal University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”, 37 Pere-
mohy av., Kyiv, 03056, Ukraine; E-mail: [email protected]
Klyushin, D.A., Taras Shevchenko National University of Kyiv, Prospekt
Glushkova 4D, 03680, Kyiv, Ukraine; E-mail: [email protected]
Kondratenko, Yuriy P., Petro Mohyla Black Sea National University, Intel-
ligent Information Systems Dept., Mykolayiv, 54003, Ukraine;
E-mail: [email protected]
Koshevoy, Nicolay, Department of Intelligent Visual Systems and Engi-
neering of Quality, National Aerospace University “Kharkov Avia-
tion Institute”, st. Chkalov, 17, Kharkov, Ukraine, 61070, E-mail:
[email protected], ORCID: http:// orcid.org/0000-0001-9465-4467
Kozlov, Oleksiy, Petro Mohyla Black Sea National University, 10 68th
Desantnykiv st., Mykolayiv, 54003, Ukraine; E-mail: [email protected]
Kuleshyn, V.V., V.M. Glushkov Institute of Cybernetics of National Academy
of Sciences of Ukraine, Kyiv, Ukraine; E-mail: [email protected]
Lebedev, D.O., Department of Computerized Control Systems, National
Shipbuilding University, Geroev Ukraine ave. 9, 54025, Mykolayiv, Ukraine;
E-mail: [email protected]
Lyashko, S.I., Taras Shevchenko National University of Kyiv, Prospekt
Glushkova 4D 03680, Kyiv, Ukraine; E-mail: [email protected]
Melnychuk, S., Space Research Institute, National Academy of Sciences of
Ukraine and State Space Agency of Ukraine, 40, b. 4/1 Glushkov av., Kyiv,
03187, Ukraine; E-mail: [email protected]
Miliavskyi, Yurii, Institute of Applied System Analysis of National Tech-
nical University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”, 37
Peremohy av., Kyiv, 03056, Ukraine; E-mail: [email protected]
Mordukhovich, Boris S., Department of Mathematics, Wayne State Univer-
sity, Detroit, Michigan 48202, USA; E-mail: [email protected]
Pysarenko, Andrii, Department of Information Systems and Technologies,
National Technical University of Ukraine “Igor Sikorsky Kyiv Polytechnic
Institute”, Kyiv, Ukraine; E-mail: [email protected]
 List of Contributors xxxiii

Romanenko, Viktor, Institute of Applied System Analysis of National Techni-

cal University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”, 37 Pere-
mohy av., Kyiv, 03056, Ukraine; E-mail: [email protected]
Roshanineshat, Arash, The University of Arizona, Electrical & Computer
Engineering Dept., Tucson, AZ, 85721, USA;
E-mail: [email protected]
Salnikov, N., Space Research Institute, National Academy of Sciences of
Ukraine and State Space Agency of Ukraine, 40, b. 4/1 Glushkov av., Kyiv,
03187, Ukraine; E-mail: [email protected]
Semenov, V., Faculty of Computer Science and Cybernetics, Taras
Shevchenko National University of Kyiv, Kyiv, Ukraine;
E-mail: [email protected]
Simon, Dan, Cleveland State University, Electrical Engineering and Com-
puter Science Dept., Cleveland, OH, 44115, USA;
E-mail: [email protected]
Skakodub, Oleksandr, Petro Mohyla Black Sea National University, 10 68th
Desantnykiv st., Mykolayiv, 54003, Ukraine;
E-mail: [email protected]
Solesvik, M., Western Norway University of Applied Sciences, Inndalsveien
28, 5063 Bergen, Norway; E-mail: [email protected]
Solovchuk, K., Poltava Scientific Research Forensic Center of the
MIA of Ukraine, Rybalskiy lane, 8, Poltava, 36011, Ukraine; E-mail:
[email protected]
Timchenko, V.L., Department of Computerized Control Systems, National
Shipbuilding University, Geroev Ukraine ave. 9, 54025, Mykolayiv, Ukraine;
E-mail: [email protected]
Tymoshenko, A.A., Taras Shevchenko National University of Kyiv, Prospekt
Glushkova 4D, 03680, Kyiv, Ukraine; E-mail: [email protected]
Vedel, Y., Faculty of Computer Science and Cybernetics, Taras Shevchenko
National University of Kyiv, Kyiv, Ukraine; E-mail: [email protected]
Vyshenskyy, V.I., V.M. Glushkov Institute of Cybernetics of National
Academy of Sciences of Ukraine, Kyiv, Ukraine; E-mail: [email protected]
Yugay, L.P., Dr.Sc.(Math), Professor, Department of Natural Sciences, Uzbek
State University of Physical Culture and Sport, Chirchik City, Republic of
Uzbekistan, Tel: +998-90-187-1478(m); E-mail: [email protected]
xxxiv List of Contributors

Zbrutsky, Alexander, Department of Aircraft Control Systems, National

Technical University of Ukraine “Igor Sikorsky Kyiv Polytechnic Insti-
tute”, Str. Botkin, 1, Kyiv, Ukraine, 03056, E-mail: [email protected],
ORCID: http://orcid.org/0000-0002-2206-7148
Zhiteckii, L., International Research and Training Center for Information
Technologies and Systems of the National Academy of Science of Ukraine
and Ministry of Education and Sciences of Ukraine, Acad. Glushkova av., 40,
Kyiv, 03187, Ukraine; E-mail: [email protected]
 List of Notations and Abbreviations

ν̄0 linear speed of point O0 relative to the origin of the

inertial coordinate system
Ω̄ = angular velocity of the CXc Yc Zc coordinate system
(Ωx Ωy Ωz )T relative to the AXYZ inertial coordinate system
Ā(t), B̄(t), C̄(t) linearized matrices AX , BX and CX
F̄v the main vector of external forces acting on the v-th
point of the system in the inertial coordinate system
M̄ M the main vector of the moment of external forces and
forces in the hinge acting on the manipulator, relative
to the origin of the inertial coordinate system
M̄0M the main vector of the moment of forces in the hinge
acting on the manipulator, relative to the point O1
p̄o1i radius vector of the ith elementary particle of the
AMR platform relative to the point O1
p̄o1j radius vector of the j th elementary particle of the
manipulator relative to the point O1
p̄m
oc radius vector of the center of mass of the manipulator
relative to the point O0
p̄ol.M radius vector of the center of mass of the manipulator
of point M relative to O1
p̄oo1 radius vector of point O1 relative to the center of mass
of the platform of point C
r̄C radius vector of point C relative to the origin of the
inertial coordinate system
v̄j linear speed of the j th point of the manipulator
relative to the origin of the inertial coordinate system
vEF engine failure velocity
V1 decision-making velocity
w̄1 = angular velocity of the O1 X1 Y1 Z1 coordinate system
(w1x w1y w1z )T relative to the CXc Yc Zc coordinate system

xxxv
xxxvi List of Notations and Abbreviations

d˜0
dt operator of local derivative in the CXc Yc Zc coordinate
system
d˜1
dt local derivative operator in the O1 X1 Y1 Z1 coordinate
system
d
dt operator of the derivative in the inertial AXY Z
coordinate system
α rudder angle
η parametric noise
ω angular speed
ψ roll angle
θ pitch angle
ε error value
ϕ yaw (course) angle
m mass of the system
mi mass of the ith elementary particle of the AMR
platform
mj mass of the j th elementary particle of the manipulator
mM mass of the manipulator
tsij –i th switching moments for j th state coordinates
A damping coefficient matrix
AX matrix parameters that depends on the coordinates
(6×6)
Am , Bm , Cm matrices of the reference mathematical model
BX matrix of control parameters that depends on the
coordinates (6×6)
CX matrix of coefficients of perturbations that depends on
the coordinates (6×6)
E error vector
F vector of external disturbances (6×1)
Fcont measured and controlled disturbances
g gravity acceleration vector (6×1)
I unit matrix
Q, G1 , G2 matrices of weighting coefficients
R space of variables
S matrices of sensitivity functions
U control vector (6×1)
Umax maximum control constraints
Ucor corrective robust control
Uopt optimal control
 List of Notations and Abbreviations xxxvii

Wcor (p) matrix transfer function of robust correction

Wo (p) matrix transfer function of the physical object
Wopt (p) matrix transfer function of the optimal mathematical
model
Wsens (p) matrix transfer function of the sensors
X state coordinate vector of moving object (6×1)
X0 , Xk vectors of initial and final conditions
f (t) external disturbance
i = 0,. . . , m order of the derivatives of state coordinate X
J optimality functional
j = 1,. . . , n order of mathematical model’s moving object
L Lagrange function
m maximum order of the positive or negative derivatives
determined by control constraints
N quadrotor UAV rotor thrust
Q quadratic functional
Tj control end time for j th state coordinates
U quadrotor UAV controls
Vw wind speed
Wf (p) transfer function of irregular waves impact
ACO Ant colony optimization
AFM Aircraft flight manual
AI Artificial intelligence
AMR Autonomous mobile robot
ASDA Accelerate-stop distance available
ASD Accelerate-stop distance
EIV Errors-in-variables
FC Fuzzy controller
FS Fuzzy system
I.I.D. Independent and identically-distributed
ICCRSU Initial conditions and control requirements setting unit
LT Linguistic terms
M Manipulator
MF Membership functions
NP Target point
Obs Obstacle
QUAV Quadrotor unmanned aerial vehicle
RB Rule base
SCU Switch control unit
xxxviii List of Notations and Abbreviations

SNR Signal-to-noise ratio

SP Start position
SVD Singular value decomposition
SW Switching key
 Part I
Advances in Theoretical Research of Control Systems
 1
Control of Moving Object Groups
in a Conflict Situation

A. Chikrii

Glushkov Institute of Cybernetics of the National Academy of Sciences

of Ukraine, 40 Glushkov Avenue, 03187, Kyiv, Ukraine
E-mail: [email protected]

Abstract
This chapter is devoted to an overview of the results concerning the differen-
tial games of pursuit–evasion with many participants on the opposing sides.
According to L. S. Pontryagin, such problems are the pinnacle of the theory of
conflict-controlled processes. The basic method of research is the method
of resolving functions. Here the problems with phase constraints, problems
of group and successive pursuit as well as the problems of interaction of
groups of participants are considered. Sufficient and sometimes the necessary
conditions for the completion of the game in a finite time are given. In
this case, quasi and stroboscopic strategies are used. The results are used
to solve several model examples from the book of R. Isaacs illustrating game
situations.1

Keywords: Differential game, guaranteed result, Pontryagin’s first direct

method, resolving function, extremal aiming rule, set-valued mapping and its
selections, stroboscopic strategy, group pursuit, state constraints, successive
pursuit.

1
This work was partially supported by National Research Foundation of Ukraine.
Grant No 2020.02/ 0121

3
4 Control of Moving Object Groups in a Conflict Situation

1.1 Introduction
Several effective mathematical methods have been developed in the theory of
dynamic games (differential games, conflict-controlled processes). In some
cases, they provide optimal results. From a practical point of view, the concept
of a guaranteed result is more appropriate. The game problems of pursuit–
evasion are the most interesting.
A survey of results in the theory of evasion is contained in [1]. The works
[2–6] provide an idea of the methods of pursuit. The last one describes the
method of resolving functions. This method theoretically substantiates the
rule of parallel pursuit and a more general method of pursuit along a beam,
well known to engineers engaged in rocket and space technology [7].
The mathematical apparatus for substantiating various schemes of the
method of resolving functions are the methods of applied nonlinear analysis,
in particular, the technique of set-valued mappings and their selections [8, 9].
The creation and development of the method [6] were stimulated by a special
‘minmaxmin’ function, introduced by the author in the theory of evasion,
and the chapter [10], which laid the foundation for group pursuit problems
[6, 10–13]. Subsequently, this method was used to study the traveling sales-
man problems, namely, the problems of successive pursuit [6–14], and the
problems with state constraints [2, 6, 15].
The method of resolving functions is based on the natural accumulative
principle. Reaching in the sum a certain value of the total winnings means the
game termination, that is, the fact that the trajectory of the conflict-controlled
process has hit a given terminal set.
The versatility of the method makes it possible to cover in a unified
scheme a wide range of dynamic processes which is proceeding in conditions
of conflict and uncertainty. These include, first of all, functional-differential
systems, stochastic processes, as well as processes with partial and fractional
derivatives [16–23].

1.2 Function ω(n, ν), Encirclement by Pshenihnyi,

Scheme of the Method of Resolving Functions

In [24] the following function was introduced:

ω(n, ν) = min max min |(pi , v)| , pi ∈ Rn , v ∈ Rn ,

pi =1 v=1 i=1, ..., ν

n ≥ 2, ν ≥ 2,
 1.2 Function ω(n, ν), Encirclement by Pshenihnyi 5

Because of the importance of this function, we analyze in detail the case of

‘simple motions’.
The dynamics of a group of ν pursuers and single evader is described by
the equations, respectively,
ẋi = ui , ui  ≤ 1, i = 1, ..., ν, xi (0) = x0i , xi ∈ Rn , (1.1)
ẏ = v, v ≤ a, y(0) = y , y ∈ R .
0 n

That the evader is captured by the ith pursuer means that xi = y. It was shown
that under the condition ω(n, ν) · a > 1 the straight motion of the evader in a
suitable direction provides avoiding meeting with the pursuers for any initial
states.
The next important step after the introduction of the function ω(n, ν)
was made by B.N. Pshenichnyi [10]. Let us put a = 1 and analyze the
problem from the point of view of a group of pursuers. In the expression
for the function ω(n, ν) we set
x0i − y 0
pi =  
x0 − y 0  , i = 1, ..., ν,
i

and remove the outer minimum and the modulus.

One can easily see that the condition max min (pi , v) > 0 is equivalent
v=1 i=1,...,ν
to the inclusion
y 0 ∈ intco{x01 , ..., x0ν }, (1.2)
This inclusion shows that at the initial moment the position of the evader
belongs to the interior of the convex hull spanned over the initial positions
of the evaders. The condition (1.2) is usually referred to as the condition for
encirclement by Pshenichnyi [10].
Pshenichnyi’s theorem. Let in the group pursuit–evasion problem (1.1)
a = 1. Then, for the group of pursuers to catch the evader in a finite time,
it is necessary and sufficient that the initial positions of the players satisfy
inclusion (1.2).
The process of group pursuit of a surrounded evader is realized using the
parallel approach strategy [7]. If the initial position of the evader lies on the
boundary or outside the convex hull spanned over the initial positions of the
pursuers, then there always exists a hyper-plane, generally speaking, loosely
separating the evader from the pursuers of the group, and the motion along the
normal to this hyper-plane in the corresponding direction insures the evader
avoiding meeting with each pursuer from the group.
6 Control of Moving Object Groups in a Conflict Situation

The method of resolving functions [25] proved its efficiency in solving

problems of group and successive pursuit. Now we outline its basic scheme
in the one-on-one case. To encompass wider classes of functional-differential
systems in a unified scheme, we consider a general representation of the
solution.
Let a conflict-controlled process be described by a generalized quasi-
linear system [25],
 t
z(t) = g(t) + Ω(t, τ ) φ(u (τ ), v (τ )) dτ, t ≥ 0, (1.3)
0

where z(t) ∈ Rn , the function g(t), g : R+ → Rn , R+ = {t : t ≥ 0}, is

measurable and bounded at t>0, matrix function Ω(t, τ ), t ≥ τ ≥ 0, is t
measurable and summable in τ for each t ∈ R+ . The control block is given
by function ϕ(u, v), ϕ : U × V → Rn , which is jointly continuous on the
direct product of nonempty sets U and V from Rn . Controls of the players
u(τ ), u : R+ → U, v(τ ), v : R+ → V, are measurable functions of time.
We assume that the terminal set M ∗ has a cylindrical form

M ∗ = M 0 + M, (1.4)

where M 0 is a linear subspace from Rn , M ∈ K(L), L is the orthogonal

complement to M 0 in Rn , K(L) is a set of nonempty compact sets from L.
We take the side of the first player (the pursuer) and define his strategies
under the assumption that the evader applies an arbitrary measurable control.
Let the game (1.3), (1.4) is evolving on the interval [0, T]. The pursuer’s
control at the moment t is selected based on the information about g(T ) and
vt (·), i.e. either u(t) = u(g(T ), vt (·)) ∈ U, where vt (·) = {v (s) : s ∈
[0, t]} is the evader’s control prehistory, or u(t) = u(g(T ), v(t)) ∈ U.
Control u(t) = u(g(T ), vt (·)) realizes quasi-strategy [5], and the
counter-control–counter-strategy u(t) = u(g(T ), v(t)), which is prescribed
by a stroboscopic strategy of O. Hayek [26].
Let π be the orthogonal projector acting from Rn to L. We set ϕ(U, v) =
{ϕ(u, v) : u ∈ U }, and consider the set-valued mappings

W (t, τ, v) = π Ω (t, τ ) ϕ (U, v), W (t, τ ) = W (t, τ, v)
v∈V

on the sets Δ × V and Δ, Δ = {(t, τ ) : 0 ≤ τ ≤ t < ∞}.

 1.2 Function ω(n, ν), Encirclement by Pshenihnyi 7

Pontryagin’s condition. The set-valued mapping W (t, τ ) is closed-valued

and has nonempty direct images on the set Δ.
Given the properties of parameters of the process (1.3), (1.4), the mapping
ϕ (U, v), v ∈ V, is continuous in the Hausdorff metric [9], the mapping
W (t, τ, v) is measurable in τ, τ ∈ [0, t], and closed in v, v ∈ V .
Therefore, the set-valued mapping W (t, τ ) is measurable in τ ∈ [0, t]. From
Pontryagin’s condition and the measurable choice theorem [9], it follows
that for any t > 0 there exists at least one τ measurable selection γ(t, τ ),
γ(t, τ ) ∈ W (t, τ ), (t, τ ) ∈ Δ.
Let us denote
 t
ξ(t, g(t), γ(t, ·)) = π g (t) + γ(t, τ ) dτ.
0
We introduce into consideration the set-valued mapping
Q(t, τ, v) == sup{α ≥ 0 : [W (t, τ, v) − γ (t, τ )] (1.5)

α [M − ξ (t, g(t), γ(t, ·))] = φ},

Q : Δ × V → 2R+ .
Its support function in the direction +1 is
α(t, τ, v) = sup{α ≥ 0 : α ∈ Q (t, τ, v)}, (t, τ ) ∈ Δ, v ∈ V. (1.6)
We call this function the resolving function [6]:
αX (p) = sup{α ≥ 0 : αp ∈ X}, p ∈ Rn , 0 ∈ X, X = X̄.
Since Pontryagin’s condition is satisfied, then the set-valued mapping
Q(t, τ, v) has nonempty closed direct images on the set Δ × V . In the
case ξ(t, g(t), γ(t, ·)) ∈ M we have Q(t, τ, v) = [0, ∞), and, therefore,
α(t, τ, v) = + ∞ for τ ∈ [0, t], v ∈ V. This corresponds to Pontryagin’s
first direct method [6]. Using the theorems on the characterization and the
inverse image [9], it can be shown that the set-valued mapping Q(t, τ, v) is
L × B-measurable for the set (τ, v), τ ∈ [0, t], v ∈ V, and the resolving
function α(t, τ, v) is joint L × B-measurable in (τ, v), by the support
function theorem [9].
Let V∗ be the set of measurable functions with values in V. We consider the
set
  t 
T (g(·), γ(·, ·)) = t > 0 : inf α(t, τ, v (τ )) dτ ≥ 1 . (1.7)
v (·)∈V∗ 0
8 Control of Moving Object Groups in a Conflict Situation

The function α(t, τ, v) is L × B-measurable, consequently, it is

super-positionally measurable [9, 27] and the integral in (2.7) has sense.
If the inequality in (1.7) does not hold for all t > 0, then we set
T (g(·), γ(·, ·)) = ∅.
Theorem 1.1 Suppose that Pontryagin’s condition holds for the conflict-
controlled process (1.3), (1.4), the set M is convex and for function g(·) and
some selection γ(·, ·)

T ∈ T (g(·), γ(·, ·)) = ∅.

Then the trajectory of the process (1.3) can be brought by the pursuer to the
terminal set (1.4) at a time T using a suitable quasi-strategy.
Moreover, if the mapping Q(t, τ, v) is convex-valued, then the process
of approach can be realized in the class of stroboscopic strategies [25].
There are several modifications of the method of resolving functions
related to various aspects of the game problem. These are a scheme for
bringing the trajectory to a given point of the set M [6], a variant of the
pursuit method when the pursuer employs only stroboscopic strategies [25],
a case when the function α(t, τ, v) is t- independent that makes it feasible
to terminate the game before the estimated time at the account of the evader’s
mistakes [6], as well as the functional form of Pontryagin’s first direct method
[25, 27]. Various forms of the pursuer advantage–Pontryagin’s condition,
bring essential changes to the method scheme. For oscillatory processes, the
Pontryagin integral condition is sometimes used [6]. For the problems of soft
meeting (the instantaneous coincidence of the players’ geometric coordinates
and velocities), an effective result is achieved by using the principle of time
dilation [28], which is based on the modified Pontryagin’s condition.
The method of resolving functions makes it feasible to study in a unified
scheme the group pursuit problems, the problems with phase constraints as
the problems with static “recumbent” pursuers [6, 15], the game problems
of successive approach—traveling salesman problems [6, 14], the pursuit–
evasion problems under the interaction of groups [6], the problem of multiple
captures [11, 12], and the problem of pursuit of the formation [12]. The form
of solution representation [1] allows extending the method to the nonlinear
case.
Not so long ago, the schemes for two types of upper and lower resolving
functions were created [29, 30], focused on the case when 0 ∈ / Q(t, τ, v)
and Pontryagin’s condition does not hold. Also, the matrix resolving function
was introduced. This makes it possible not only to attract the set M −
 1.3 Group Pursuit of a Moving Object 9

ξ(t, g (t), γ (t, ·)) to the intersection from (2.5) in the cone spanned on
it, but to rotate it as well [31].

1.3 Group Pursuit of a Moving Object

When a group of ν pursuers and a single evader are evolved in the group
pursuit, the following effects can be observed. Each of the pursuers from the
group cannot independently catch the evader in a finite time, but by acting
together they can achieve the desired goal [6, 11–13, 15]. Another case is
when each pursuer can individually intercept the evader, but acting together
in a group they can reach the desired goal faster.
Formally, the problem setting should contain the dynamics (1.3) and a
terminal set–the set of ν cylindrical sets Mi∗ , i = 1, ..., ν, of the form
(1.4). To emphasize the presence of several pursuers, we consider the conflict-
controlled process,
 t
zi (t) = gi (t)+ Ωi (t, τ ) ϕi (ui (τ ), v (τ )) dτ, i = 1, ..., ν, t ≥ 0, (1.8)
0

where zi (t) ∈ Rni , gi : R+ → Rni , functions gi (t) are measurable and

bounded in t > 0, matrix functions Ωi (t, τ ), t ≥ τ ≥ 0, are measurable
in t and summable in τ for each t ∈ R+ . Functions ϕi (ui , v), ϕi : Ui ×
V → Rni , are jointly continuous in their variables, and Ui , V are nonempty
compacts. The admissible controls of the players are measurable functions.
The pursuers apply quasi or stroboscopic strategies [5, 6, 26].
The terminal set has the form
Mi∗ = Mi0 + Mi , i = 1, ..., ν, (1.9)
where Mi0 are linear subspaces from Rni , and Mi ∈ K(Li ), Li are
orthogonal complements to Mi0 in Rni .
Let us denote by zi a pair (xi , y), where xi and y are the states of the ith
pursuer and the evader, respectively. Since the goal of the group of pursuers
consists in the capture of the evader by at least one of the pursuers, then vector
zi hitting the set Mi∗ means the capture of the evader by the ith pursuer. . .
Let πi be the operators of orthogonal projection from Rni to Li , and
ϕi (Ui , v) = {ϕi (ui , v) : ui ∈ Ui }, v ∈ V. Let us take into consideration the
forms of the set-valued mappings, respectively:

Wi (t, τ, v) = πi Ωi (t, τ ) ϕi (Ui , v), Wi (t, τ ) = Wi
v∈V
10 Control of Moving Object Groups in a Conflict Situation

(t, τ, v), t ≥ τ ≥ 0, i = 1, ..., v.

Pontryagin’s condition for a group of pursuers. The set-valued mappings

Wi (t, τ ), (t, τ ) ∈ Δ, i = 1, ..., ν, are closed-valued and have nonempty
direct images.
Since the set-valued mappings Wi (t, τ ), i = 1, ..., ν, are mea-
surable in τ and closed-valued, then there exist τ measurable selections
γi (t, τ ), γi (t, τ ) ∈ Wi (t, τ ), (t, τ ) ∈ Δ, analogously to the case ν = 1.
We set
 t
ξi (t, gi (t), γi (t, ·)) = πi gi (t) + γi (t, τ ) dτ, i = 1, ..., ν,
0

and introduce numerical set-valued mappings

Qi (t, τ, v) =

= {αi ≥ 0 : [Wi (t, τ, v) − γi (t, τ )] αi [Mi − ξi (t, gi (t), γi (t, ·))] = ∅},
Qi : Δ × V → 2R+ .

Below are given their support functions in the direction +1:

αi (t, τ, v) = sup{αi ≥ 0 : αi ∈ Qi (t, τ, v)}, t ≥ τ ≥ 0,

v ∈ V, i = 1, ..., ν.
Let us set

g ν (·) = (g1 (·) ... gν (·)), γ ν (·, ·) = (γ1 (·, ·)...γν (·, ·)),

and take a closer look at the set

 
t
T (g (·), γ (·, ·) =
ν ν ν
t > 0 : inf max
______ αi (t, τ, v(τ ))dτ ≥ 1 .
v(·)∈V∗ i= 1,ν 0
(1.10)
If for some index i ∈ {1, ..., ν} and time instant t > 0 inclusion
ξi (t, gi (t), γi (t, ·)) ∈ Mi , has a place, then at the time instant t the ith
pursuer can catch the evader, i.e. Qi (t, τ, v) = [0, ∞), αi (t, τ, v) = + ∞,
for all τ ∈ [0, t], v ∈ V , and the capture is realized in the class of
stroboscopic strategies in compliance with the scheme of the first direct
method [3].
In the case the inequality in (1.10) is not fulfilled, we put
T (g ν (·), γ ν (·, ·)) =φ.
ν
 1.3 Group Pursuit of a Moving Object 11

Theorem 1.2. Let for the group pursuit problem (1.8), (1.9) an analog of
Pontryagin’s condition be satisfied, the sets Mi , i = 1, ..., ν, are con-
vex and suppose that for function g ν (·) there exist measurable selections,
γ ν (·, ·), γi (t, τ ) ∈ Wi (t, τ ), (t, τ ) ∈ Δ, i = 1, ..., ν, and an integer
T ν ∈ T ν (g ν (·), γ ν (·, ·)) =φ.
Then there is a quasi-strategy of the pursuers such that at least one of the
trajectories (1.8) can be brought to the corresponding set at the time instant
T ν for any admissible control of the evader.
By the way of illustration, below is given a simple example of the group
pursuit with the same control domains of the players

żi = ui − v, zi ∈ Rni = Rn , n ≥ 2, ui  ≤ 1, v  ≤ 1, zi (0) = zi0 ,

(1.11)
and the captures described by the sets:

Mi∗ = Mi0 = Mi = {zi : zi = 0}, i = 1, ..., ν. (1.12)

Here Li = Rn , πi are identity transformation operators πi : Rn → Rn ,

described by the n-dimensional unit matrix E.
Here in the presentation of solution (1.8) gi (t) = zi0 , Ω(t, τ ) = E,
ϕi (ui , v) = ui − v, i = 1, ..., ν, see (1.8), and the Pontryagin’s condition
for a group of pursuers holds: Wi (t, τ ) = S ∗S = 0, S = {p : p < 1},
with γi (t, τ ) = 0.
Here αi (t, τ, v) = αi (zi0 , v) = max{α ≥ 0 : − αzi0 ∈ S − v}, the
resolving functions show up as the greatest positive roots of the quadratic
equations  v − αi zi0  = 1, i = 1, ..., ν, and have the form

(zi0 , v) + (zi0 , v)2 +  zi0 2 (1 − v 2 )

αi (zi0 , v) = , i = 1, ..., ν.
 zi0 2

The encirclement by Pshenichnyi [10] for the initial states z 0 = (z10 , ..., zν0 )
can be expressed in similar forms [6], namely:
1) δ(z 0 ) = min max αi (zi0 , v) > 0;
 v  ≤ 1 i=1,...,ν
 0 
zi
2) σ(z 0 ) = min max  zi0 
, v > 0;
 v  ≤ 1 i=1,...,ν
0 
z
3) 0 ∈ intco  zi0  .
i

Below is given the statement that provides necessary and sufficient conditions
for the successful termination of the group pursuit–evasion and the pursuit at
12 Control of Moving Object Groups in a Conflict Situation

the moment
T0ν = min T ν (z 0 , 0).
0 
z
Corollary 1.1. If 0 ∈ intco  zi0  , where z 0 is the initial state of the
i
process (1.11), then the group pursuit can be terminated at the time T0ν and
the following estimate is valid: T0ν ≤ δ(zν0 ) . Herewith, if t∗ = t∗ (v(·)) is the
time of control switching, i.e. a zero of the control function
 t
1 − max αi (zi0 , v (τ )) dτ,
i=1,...,ν 0

then the controls of the pursuers, realizing time T0ν on the active section
[0, t∗ ], are

ui (τ ) = v(τ ) − αi (zi0 , v(τ )) zi0 , i = 1, ..., ν, (1.13)

and on the passive one (t∗ , T0ν ] ui (τ ) = v(τ ) for indices i, satisfying the
equality
 t∗  t∗
0
αi (zi , v(τ )) dτ = max____ αi (zi0 , v (τ )) dτ,
0 i∈ 1,ν 0

and for the rest of the indices, the controls of the pursuers at the interval
(t∗ , T0ν ] can be arbitrary admissible. If
 0 
zi
0∈intco
¯ ,
 zi0 
then the escape is possible.
Corollary 1.2. Let in the group pursuit problem (1.11)

Mi0 = {0}, Mi∗ = Mi = εi S, εi > 0, i = 1, ..., ν.

Then, if
0 ∈ int co {zi0 + εi S},
then at least one of the pursuers can catch the evader in a finite time. If

0 ∈
¯ int co {zi0 + εi S},

then the escape is possible at all semi-infinite intervals of time.

 1.3 Group Pursuit of a Moving Object 13

Simple matrix. The dynamics are given by the equations

żi = ai zi + ui − v, ai < 0, zi ∈ Rn , ui  ≤ 1, v  ≤ 1, i = 1, ..., ν.

The goal of the pursuer’s group is the exact capture of the evader by at
least one of the pursuers (1.12). The result has the form of necessary and
sufficient conditions [10], which is similar to Corollary 1 and realizes a
parallel approach (1.13) on the active section, with the only difference that
the following estimate is fulfilled for the guaranteed time
 
1 a∗ · ν
T (z , 0) ≤
ν 0
ln 1 + , a∗ = − max
____ ai .
a∗ δ(z 0 ) i= 1,ν

Second-order objects.
żi1 = zi2 ,

żi2 = − a1 zi2 − a2 zi1 + ui − v, ui  ≤ 1, v  ≤ 1, i = 1, ..., ν. (1.14)

The terminal set is described by the sets

Mi∗ = Mi0 = Mi = {zi = (zi1 , zi2 ) : zi1 = 0}, i = 1, ..., ν,

where zi1 is the difference between the geometric coordinates of the pursuer
xi and the evader y, zi1 = xi − y, and zi2 is the difference of their speeds,
zi2 = ẋi − ẏ.
Because of the equivalence of the Equation (1.14) to the second-order
equations:
ẍi + a1 ẋi + a2 xi = ui ,
ÿ + a1 ẏ + a2 y = v,
the characteristic polynomial takes the form

λ2 + a1 λ + a2 = 0. (1.15)

We will restrict ourselves to the case of real roots and present the results in the
form of consequences of the general method scheme for classical examples
[13, 15].
14 Control of Moving Object Groups in a Conflict Situation

The pursuers and the evader are material points [5]

Let a1 = a2 = 0 in (1.14). Then the roots of the polynomial (3.8) are λ1 =
λ2 = 0.
 01 02
∗ ∗ zi , zi = 0,
If 0 ∈ intco {zi }, where zi = thenT ν (z 0 , 0) < + ∞.
zi02 , zi02 = 0,
Pontryagin’s test case [3]
Let a2 = 0, a1 > 0 in (1.14). Then λ1 = − a1 , λ2 = 0.
If 0 ∈ intco {zi∗ }, where zi∗ = a1 zi01 + zi02 , then T ν (z 0 , 0) < + ∞.
If the objects are of a different type, then a1 is changed for ai1 , and in the
encirclement, the condition zi∗ should be put equal to zi01 + a1i zi02 . Then the
1
result holds [13, 15].

1.4 Non-fixed Time of Game Termination

The scheme of the method of resolving functions allows, under certain
assumptions, to obtain the conditions for the game termination not only at the
moment but also no later than in a certain time, using the evader’s mistakes
[6]. Let the motion of a group of controlled objects z = (z1 , ..., zν ) in space
Rn be described by quasi-linear equations

żi = Ai zi + ϕi (ui , v), i = 1, ..., ν, (1.16)

zi ∈ Rni , n = n1 + ... + nν , Ai are square matrices of the order ni , ui ∈ Ui ,

v ∈ V, Ui , V are nonempty compact sets, functions ϕi (ui , v) are jointly
continuous in their variables.
The terminal set consists of cylindrical sets (1.9), where Mi , i = 1, ..., ν,
are convex compact sets, As usual, πi are the ortho–projector acting from Rni
___
to Li , con X = {x ∈ X : λx ∈ X ∀λ > 0}, con X is the cone closure.
Condition 1.1. For the fixed
 point z ∈ Rn ,
___
con (Mi − πi etAi zi ) πi e(t−τ )Ai ϕi (Ui , v) =φ, i = 1, ..., ν, (t, τ ) ∈
Δ, v ∈ V.
Such points are given, let us introduce the resolving functions 
αi (t, τ, zi , v) = sup {αi ≥ 0 : π e(t−τ )Ai ϕi (Ui , v) α [Mi −
πi etAi zi ] =φ},
i = 1, ..., ν, (t, τ ) ∈ Δ, v ∈ V,
 1.4 Non-fixed Time of Game Termination 15

and set
 
t
ν
T (z) = min t ≥ 0 : inf max
_____ αi (t, τ, zi , v (τ )) dτ ≥ 1 .
v(·)∈V∗ i= 1,ν 0

Theorem 1.3. Suppose that for the process (1.16), (1.9) Mi = {0}, πi Ai =
Ai πi , i = 1, ..., ν, the initial position z 0 satisfies Condition 1.1, and
T ν (z 0 ) < + ∞.
Then the group pursuit can be terminated starting from the initial state z 0
no later than in time T ν (z 0 ) in the class of stroboscopic strategies.
Let us put Ai = {0}, i = 1, ..., ν. Then
żi = ϕi (ui , v), i = 1, ..., ν. (1.17)
Now we consider the set-valued mappings
 ____
Wi (zi , v) = πi ϕi (Ui , v) con(Mi − πi zi ),
 ____
W̃i (zi , v) = co {πi ϕi (Ui , v)} con(Mi − πi zi ),
and for zi ∈ Rni \Mi∗ the sets
Wi = {zi : Wi (zi , v) = φ∀v ∈ V },
W̃i = {zi : W̃i (zi , v) = φ∀v ∈ V }.
The resolving functions have the forms

ᾱi (zi , v) = max {α ≥ 0 : πi ϕi (Ui , v)

α[Mi − πi zi ] == φ}, zi ∈ Wi , v ∈ V,


α̃i (zi , v) = max {α ≥ 0 : co [πi ϕi (Ui , v)] α[Mi − πi zi ] == φ},

zi ∈ W̃i , v ∈ V, i = 1, ...ν.
We set
 t
T̃ ν (z) = min {t ≥ 0 : inf max
_____ ᾱi (zi , v(τ ))dτ ≥ 1} .
v(·)∈V∗ i= 1,ν 0

Let us introduce the summary functions

ᾱ(z) = inf max
_____ ᾱi (zi , v), (1.18)
v∈V i= 1,ν
16 Control of Moving Object Groups in a Conflict Situation

α̃(z) = inf max

_____ α̃i (zi , v).
v∈V i= 1,ν

Theorem 1.4. Let zi0 ∈ Wi , i = 1, ..., ν, for the initial state of the process
(1.17) z 0 and ᾱ(z 0 ) > 0. Then the group pursuit problem can be solved in the
class of counter-controls no later than in time T̃ ν (z 0 ) and T̃ ν (z 0 ) ≤ ᾱ(zν 0 ) .
If, otherwise, for z 0 , zi0 ∈ W̃i , i = 1, ..., ν, and α̃(z 0 ) = 0, then the
problem of evasion from a group of pursuers can be solved under the constant
control of the evader, furnishing minimum in (1.18).
Let be

żi = ui − v, zi ∈ Rn , v ∈ V, ui ∈ ∂coV, i = 1, ..., ν. (1.19)

Here V is a compact set from Rn , ∂coV is the boundary of the set coV. The
terminal set consists of convex compact sets M1 , ..., Mν from Rn .
Now we deduce necessary and sufficient conditions, ensuring solvability
of the group pursuit problem (1.19) for given initial states.
Let us set
  ____
W i (zi , ν) = [∂coV − ν] con[Mi − zi ],
 
W i = {zi :W i (zi , ν) = φ∀ν ∈ V }, i = 1, ..., ν,
 
α i (zi , ν) = max{α ≥ 0 : [∂coV − ν] [Mi − zi ] = φ},

zi ∈W i , zi ∈
¯ Mi , v ∈ V,

 
α (z) = inf max
_____
α i (zi , v),
v∈V i= 1,ν
 
t

T̂ ν (z) = min t ≥ 0 : inf max
_____
αi (zi , v (τ )) dτ ≥ 1 .
v(·) i= 1,ν 0

We now take into consideration the support function and the support set for
an arbitrary compact set X from the space Rn , respectively.

CX (p) = max (x, p), p ∈ Rn ,

x∈X

H(X; p) = {x ∈ X : (p, x) = CX (p)}, p = 0.

 1.5 The Group Pursuit. Linear State Constraints 17

In the case the images of the mapping H(X; p) consist of single points for
any p ∈ Rn , they say that X is a strictly convex compact.
 The set X is called a compact with smooth boundary if H(X; p1 )
H(X; p2 )=φ ∀p1 , p2 , p1 = p2 , p1  = p2  = 1.
Lemma 1.1. Let for the process (1.19) the set V be a strictly convex compact

with smooth boundary. Then α (z 0 ) > 0 if and only if
⎡ ⎤

0 ∈ intco ⎣ (Mi − zi0 )⎦ . (1.20)
_______
i=1,...,ν

It should be emphasized that in the case that V is a strongly convex compact


set, the inequality α (z 0 ) > 0 leads to inclusion (1.20), but the opposite,
generally speaking, is not true. If V is a compact set with a smooth boundary,

then inclusion (1.20) implies inequality α (z 0 ) > 0,the opposite is generally
not true.
Theorem 1.5. Let for the conflict-controlled process (1.19) V be a
strictly convex compact set with a smooth boundary. Then, if 0 ∈

intco (Mi − zi0 ) , then the group pursuit can be terminated starting
______
i=1,...,ν
from the initial state z 0 no later than in the time T̂ ν (z 0 ). If, otherwise,

0∈¯ intco (Mi − zi0 ) , then starting from the initial state z 0 it is
______
i=1,...,ν
possible to avoid meeting the group of pursuers.

1.5 The Group Pursuit. Linear State Constraints

In what follows, we show that the group pursuit problem under linear state
constraints on the evader’s state can be reduced to the equivalent group pursuit
problem without state constraints with a larger number of pursuers [6, 13, 15].
Consider the conflict-controlled process (1.16)

żi = Ai zi + ϕi (ui , v), zi ∈ Rni , ui ∈ Ui , v ∈ V, zi (0) = zi0 , i = 1, ..., ν.

The terminal set consists of cylindrical sets

Mi∗ = Mi0 + Mi , i ∈ Nν = {1, ..., ν}. (1.21)

18 Control of Moving Object Groups in a Conflict Situation

Here Mi0 are linear subspaces from Rni , Mi are nonempty convex compact
sets from Li , the orthogonal complements to Mi0 in Rni , and Mi = {ai },
i ∈ Nν \Nk , k ≤ ν, and ai are vectors from Li , dim Li = 1.
Consequently, for i ∈ {k + 1, ..., ν}, dim Mi0 = ni − 1 and sets Mi∗ appear
as affine manifolds. We assume that the pursuers use quasi-strategies. We
denote by πi , the operator of orthogonal projecting from Rni to Li , i ∈ Nν .
Let us introduce the set-valued mappings
Wi (t, τ ) = πi eAi t ϕi (Ui , v), i ∈ Nν .

For i ∈ Nk , we set Wi (t) = Wi (t, v), t ≥ 0.
v∈V
Condition 1.2. Direct images of mappings Wi (t) are nonempty for all i ∈
∈ Nk , t ≥ 0.
We fix some measurable selections γi (t), γi (t) ∈ Wi (t), i ∈ Nk , and set
 t
0 tAi 0
ξi (t, zi , γi (·)) = πi e zi + γi (τ ) dτ.
0

Let us define the resolving functions for i ∈ Nk


αi (t, τ, zi0 v, γi (·)) = sup {α ≥ 0 : [Wi (t − τ, v) − γi (t − τ )]

α[Mi − ξi (t, zi0 , γi (·))] = φ}.

They enjoy all the properties provided earlier. Consider the case i ∈ Nν \Nk .
Condition 1.3. Set-valued mappings Wi (t, v), i ∈ Nν \Nk , are continuous
single-valued functions ωi (t, v).
Let us set
ξi (t, zi0 ) = πi etAi zi0 , i ∈ Nν \Nk ,
and write the resolving functions for i = k + 1, ..., ν:

(pi , ωi (t−τ, v))
0 , pi ∈ Li , pi  = 1, ai = ξi (t, zi ),
0
∗ 0 (p i , ai −ξi (t,zi ))
αi (t, τ, zi v) =
ωi (t − τ, v) + 1t , ai = ξi (t, zi0 ).
It is supposed that they are continuous in t, t > 0, γ(·) = (γ1 (·), ..., γν (·).
Let us take into consideration the function
   t
ν 0
T∗ (z , γ(·)) = min t ≥ 0 : inf max max αi (t, τ, zi0 , v (τ ),
v(·) i∈Nk 0
 1.5 The Group Pursuit. Linear State Constraints 19
 t  
γi (·)) dτ, max αi∗ (t, τ, zi0 , v (τ )) dτ ≥1 .
i∈Nν \Nk 0

If the inequality in braces is not satisfied, then we set T∗ν (z 0 , γ(·)) = + ∞.

Theorem 1.6. Let the parameters of the conflict-controlled process (4.1),
(5.1) meets Conditions 1.2 and 1.3 and let for the initial state z 0 and some
measurable selections

γ(t), γi (t) ∈ Wi (t), i = 1, ..., k, T∗ν (z 0 , γ(·)) < + ∞.

Then, the trajectory of the process (1.16) can be brought from the initial state
z 0 to the corresponding set Mi∗ at the moment T∗ν (z 0 , γ(·)) for at least one
i, i ∈ Nν , using quasi-strategies.
By way of illustration, let us study the linear problem of control of a group
of pursuers when the evader cannot leave some open polyhedral set.
The dynamics of the pursuers and the evader have the forms, respectively:

ẋi = Ci xi + ui , xi ∈ Rri , ui ∈ Ui , i = 1, ..., k, (1.22)

ẏ = B y + v, y ∈ Rs , v ∈ V,
Ci are quadratic matrices of order ri , and B is a matrix of the order s, Ui , V
are nonempty compact sets. The motion of the evader is bounded by the
constraints

G = {y ∈ Rs : (pi , y) < li , pi ∈ Rs , pi  = 1, i = k + 1, ..., ν} = ∅.

(1.23)
We see that the terminal set consists of cylindrical sets Mi∗ , i = 1, ..., k,
from the spaces Rni = Rri × Rs .
We say that at some moment the pursuit is completed if at this moment
at least one of the pursuers catches the evader, i.e. for some i ∈ Nk zi =
(xi , y) ∈ Mi∗ , or the evader is forced to violate the state constraints, that is,
(pi , y) = li for some i ∈ Nν \Nk .
For i ∈ Nν \Nk we set

zi = y, Ai = B, ϕi (u, v) = v,

Mi0 = {zi : (pi , zi ) = 0}, Mi = {ai : ai = li pi }, Mi∗ = {zi : (pi , zi ) = li }.

20 Control of Moving Object Groups in a Conflict Situation

Thus, the group pursuit problem (1.22) under the state constraints (1.23)
is reduced to a group pursuit problem (1.16), (1.21) without constraints.
The above-outlined scheme can be used to study the game problem with
arbitrary state constraints being convex sets [6].
Let us analyze in detail group pursuit for ‘simple motions’
ẋi = ui , ui  ≤ 1, xi ∈ Rs , s ≥ 2, i = 1, ..., k, ẏ = v v  ≤ 1, y ∈ Rs ,
(1.24)
under the state constraints (1.23).
The group pursuit terminates if xi − y  ≤ εi , εi ≥ 0, for some i ∈ Nk , or
the evader hits the boundary of set G, i.e. for some i ∈ Nν \Nk the equality,
(pi , y) = li takes place. The game problem (1.24), (1.23) is easily reduced
to a differential group pursuit game. Let us put zi = xi − y. i ∈ Nk . Then
żi = ui − v, ui  ≤ 1, v  ≤ 1, Mi∗ = {zi : zi  ≤ εi }, i = 1, ..., k.
(1.25)
We set zi = y., i ∈ Nν \Nk . Then we have
żi = v, v  ≤ 1, Mi∗ = {zi : (pi , zi ) = li }, i = k + 1, ..., ν. (1.26)
It is evident that the analog of Pontryagin’s condition for a group is
fulfilled, the selections γi (t) ≡ 0, i ∈ Nk , t ≥ 0. Hence, the functions
ξi (t, zi0 , 0) = zi0 , ξi (t, zi0 ) = zi0 , i ∈ Nν \Nk , ωi (t, v) = v, and the
resolving functions take the forms
αi (t, τ, zi0 , v, 0) = ρi (zi0 , v) =

 2
(v, zi ) + εi + ((v, zi0 ) + εi )2 + (zi0  − ε2i )(1 − v2 )
0
=  2 , i ∈ Nk ,
z 0  − ε2
i i
(pi , v)
αi∗ (t, τ, zi0 , v) = ρi (zi0 , v) = , z 0 = ai = li pi , i ∈ Nν \Nk .
li − (pi , zi0 ) i
The termination time of the group pursuit game (1.25), (1.26) is defined as
the smallest positive root of the equation
 t
inf max ρi (zi0 , v(τ )) dτ = 1. (1.27)
v(·) i∈Nν 0

In the sequel, we provide results concerning problems of group pursuit–

evasion in various forms, in their number, in the form of encirclement about
the initial states [6, 12].
 1.5 The Group Pursuit. Linear State Constraints 21

Let vectors b1 , ..., bν , bi  = 1, belong to space Rs . They form a positive

basis if for each vector z ∈ Rs there exist positive numbers σi , i = 1, ..., ν,
ν
such that z = i=1 σi bi . Let us put

βi (bi , v) = (bi , v) + (bi , v)2 + 1 − v 2 , i = 1, ..., k,

βi (bi , v) = (bi , v), i = k + 1, ..., ν,

β(b1 , ..., bν ) = min max
_____ βi (bi , v).
 v ≤1 i= 1,ν

Lemma 1.2. The following statements are equivalent:

1) vectors b1 , ..., bν form a positive basis;
2) vectors b1 , ..., bν form a Carathéodory set, i.e. 0 ∈ int co{b1 , ..., bν };
_____
3) con {b1 , ..., bν } = Rs ;
4) min max _____(bi , v) > 0;
 v =1 i= 1,ν
5) β(b1 , ..., bν ) > 0.

Theorem 1.7. Let z 0 = (z10 , ..., zν0 ) be the initial state of the process (1.25),
(1.26). If there is a number i ∈ Nk , such that εi > 0, or k ≥ s, and

0 ∈ int {co [z10 + ε1 S, ..., zk0 + εk S] + con (pk+1 , ..., pν )}, (1.28)

then there exists a finite root of equation (1.27).

Corollary 1.3. Let z 0 be the initial state of the process (1.25), (1.26). Then, if
0∈
¯ int co{z10 + ε1 S, ..., zk0 + εk S, pk+1 , ..., pν }, then the escape is possible
on set Ḡ.
Corollary 1.4. Let for the process (1.25), (1.26) the set G be a polyhedron (a
bounded set), and one of the following conditions hold:
1) for some i ∈ Nk εi > 0;
2) k ≥ s.
Then the group pursuit problem (1.25), (1.26) with state constraint Ḡ can be
solved in a finite time for any initial states.
Remark. If for the process (1.25), (1.26) εi = 0 for i ∈ Nk , k < s, then even
in the case the inclusion 0 ∈ int co {z10 , ..., zk0 , pk+1 , ..., pν } is fulfilled,
the escape on a set Ḡ is possible. This follows from the results from [6].
Below we analyze some classical problems with state constraints from the
book [2], which can be solved based on Theorem 1.7 and its corollaries [6].
22 Control of Moving Object Groups in a Conflict Situation

Lion versus man [2]. Let k = 1, ε1 > 0, the pursuer and the evader move
according to equations (1.24), ν − 1 > s, set G be bounded and given by
inequalities (1.23).
Then con{p2 , ..., pv } = Rs and inclusion (1.28) is satisfied for all initial
positions. Hence, capture is always possible. What is more, the result holds
in the case G be an arbitrary convex compact.
Rat cornered [2]. Let motion of a cat and a rat be described by the equations
(1.24), respectively, with k = 1, ε1 > 0, a set G be a convex cone K in
space Rs . Then the barrier cone of a set G, KG = − K ∗ , where K ∗ is a cone
conjugate to K, given the form of the support function

0, p ∈ −K ∗ ,
CK (p) = ¯ − K ∗.
+∞, p∈
Consequently, if there are vectors pi , i = 2, ..., ν, pi ∈ − K ∗ ,pi  = 1,
such that y 0 ∈ int{x0 + ε1 S+ +con [p2 , ..., pν ]}, then the capture can be
performed in a finite time.
Patrolling the corridor [2]. Let k = 1, s = 2, ε1 > 0. State constraints are
two parallel straight lines on the plane having the normals p2 , p3 . Without
loss of generality, we can assume that these straight lines are parallel to the
abscissa axis. We see that p2 and p3 are orthogonal to the latter. Therefore,
the condition, which is sufficient for the capture (1.28), takes the form | x01 −
y10 | < ε1 , where x01 , y10 are the first coordinates of the initial states of the
pursuer and the evader, respectively.
Death line game [2]. Let k = 1, ε1 > 0, s = 2. The game takes place on a
half-plane–below the abscissa axis, which is the line of death for the evader,
and the normal p2 coincides in direction with the axis of ordinates. The
inequalities | y10 − x01 | < ε1 , y20 >x02 separate the initial positions of the pur-
suer and the evader, starting from which the capture is possible in a finite time.
‘Simple’ group pursuit. Let k = ν ≥ s, εi ≥ 0. Then sufficient and
necessary condition for capture in the group pursuit problem (1.25) is
0 ∈ int co {z10 + ε1 S, ..., zk0 + εk S}..

1.6 Principle of Shortest Broken Line in Successive

Pursuit
Now we proceed to the analysis of the problem of successive capturing of a
group of evaders by a single pursuer. Various schemes have been developed
 1.6 Principle of Shortest Broken Line in Successive Pursuit 23

with a fixed and non-fixed moment of capture of each of the evaders. The
proposed schemes [6, 14] are especially effective in the case of ‘simple
motions’ with the beforehand programmed choice of the order of capture.
Then the pursuit strategy is the parallel approach and, consequently, the total
capture time depends only on the evader’s controls. In so doing, extremum
of the functional is attained on constant controls of evaders, and the infinite-
dimensional problem of maximizing the total pursuit time is reduced to a
finite-dimensional conditional optimization problem.
Let a conflict-controlled process be described by the quasi-linear
equation
żj = Aj zj + ϕj (u, vj ), z j ∈ Rnj , u ∈ U, vj ∈ Vj , j = 1, ..., μ, (1.29)
Aj are square matrices of the order nj , u, vj are the control parameters
of the pursuer and evaders belonging to the nonempty compacts U and Vj ,
respectively, functions ϕj (u, vj ) are jointly continuous in their variables.
The terminal set consists of cylindrical sets
____
Mj∗ = Mj0 + Mj , Mj∗ ⊂ Rnj , j = 1, μ, (1.30)
where Mj0 are linear subspaces from Rnj , and Mj are convex compact sets
from the orthogonal complements Lj to Mj0 in Rnj .
The
____
goal of the pursuer is to bring in turn all the trajectories zj (t),
j = 1, μ, to the corresponding sets Mj∗ in the least total time. The pursuer
is faced with a double problem, dealing with control and enumerating, which
can in no way be divided and studied separately. Let us denote by zj a pair
(x, yj ), where x is the pursuer state and yj -the evaders. Then inclusion
zj ∈Mj∗ signifies the capture of the jth evader by the pursuer. We suppose
that the pursuer applies quasi-strategies.
We state that the successive pursuit can be completed, starting from the
initial state, z 0 = (z10 , ..., zμ0 ) no later than at the time T, if there is a quasi-
strategy of the pursuer such that for ____
any controls of evaders vT (·) there ____
exist
∗
moments tj , 0 < tj ≤ T, j = 1, μ, such zj (tj ) ∈ Mj for all j = 1, μ .
It should be noted that the values of the function vj (t) are used only on time
intervals [0, tj ]. In the case the order of bringing trajectories zj (t) to the
corresponding sets Mj∗ is chosen beforehand, i.e. is fixed at the initial moment
and then does not change, the number of possible options is μ_!
We denote by Λ the set of all possible orders for capturing the evaders.
Let us assume that at the initial moment the pursuer selects some order
l, l ∈ Λ, and then sticks it. Without loss of generality, we determine by
l = {1, 2, ..., μ} the order of bringing the trajectories to corresponding sets.
24 Control of Moving Object Groups in a Conflict Situation

Let us embark on minimization of the total time. We denote by πj be the

operator of orthogonal projecting from Rnj to Lj . We set

W j (t, vj ) = πj eAj t ϕj (U, vj ), W j (t) = W j (t, vj ),
vj ∈Vj

j = 1, ..., μ, t ≥ 0.
Condition 1.4. The direct images of set-valued mappings W j (t) are not
empty and closed for all j = 1, ..., μ, t ≥ 0.
Since the mappings W j (t) are measurable and closed-valued, then there
exist measurable selections γj (t), γj (t) ∈ Wj (t), t ≥ 0, j = 1, ..., μ [9].
We set  t
Aj t
ξj (t, zj , γj (·)) = πj e zj + γj (τ ) dτ.
0
Then the resolving functions take the forms
αj (t, τ, zj , vj , γj (·)) = sup {α ≥ 0 : [W j (t − τ, vj ) − γj (t −

τ )] α[Mj − ξj (t, zj , γj (·))]=φ}, j = 1, ..., μ, t ≥ τ ≥ 0.
We see that they enjoy all the above-listed properties.
Now we determine recursively the moments of trajectories (1.29) hitting the
corresponding sets (1.30) in order l. The following relations are true:
tj = tj (zj (tj−1 ), vj (·)) = tj (z10 , ..., zj0 , v1 (·), ..., vj (·)),
Here
 
tj
tj = min t ≥ tj−1 : αj (t, τ, zj (tj−1 ), vj (τ ), γj (·)) dτ ≥ 1 ,
tj−1

j = 1, ..., μ.
It is evident that the total time of capturing all the evaders coincides with the
moment of capturing the last evader and, consequently, is equal to tμ .
Let us denote z 0 = (z10 , ..., zμ0 ), v(·) = (v1 (·), . . . , vμ (·)) , and by
Tμl (z 0 , v(·)) = tμ (z10 , ..., zμ0 , v1 (·), ..., vμ (·))
–the total capture time.
This time depends on the order of capturing the evaders and its optimiza-
tion in the case of the programmed order of capture consists in maximizing
in v(·) and minimizing concerning the order, i.e.
Tμ (z 0 , γ(·)) = min sup Tμl (z 0 , v(·)),
l∈Λ v(·)
 1.6 Principle of Shortest Broken Line in Successive Pursuit 25

γ(·) = (γ1 (·), ..., γμ (·)) are the fixed forehand selections.
Theorem 1.8 Let Condition 1.4 be satisfied, and there exist measurable
selections of set-valued mappings γj (t) ∈ W j (t), t ≥ 0, j = 1, ..., μ,
such thatTμ (z 0 , γ(·)) < + ∞.
Then, starting from the initial state z 0 , the successive pursuit can be
terminated at the time instant Tμ (z 0 , γ(·)), using quasi-strategies.
To apply the modified method scheme with non-fixed termination time to
solving the successive pursuit problem using stroboscopic strategies requires
significantly more burdensome constraints on the process parameters. It is
suited to the case of ‘simple motions’. For the sake of simplicity and geo-
metric descriptiveness, we restrict ourselves to the case of plane motions.
Let be
żj = u − vj , zj ∈ R2 , zj (0) = zj0 ,
u ≤ 1,  vj  ≤ β < 1, j = 1, ..., μ, (1.31)
zj = x − yj , ẋ = u, ẏj = vj .
It is easy to observe that each of the pursuers has an advantage over each of
the evaders, otherwise, the problem has no sense. The goal of the pursuer is to
bring in turn all trajectories zj (t) to the origin in a finite time, i.e. Mj∗ :zj = 0.
As before, it is assumed that the system with the lower index has priority in
the order of bringing. We denote

tj = min {t > 0 : zj (t) = 0} (1.32)

the moment of the first hitting the set Mj∗ . by the trajectory zj (t).
We have

Mj0 = Mj = {0}, Lj = R2 , πj = I, Aj = {0},

ϕj (u, vj ) = u − vj , j = 1, ..., μ.
Then

W j (t, vj ) = S − vj , W j (t) = (1 − β) S, j = 1, ..., μ, t ≥ 0.

Since β < 1, then Condition 1.4 is fulfilled. We select γj (t) ≡ 0, j =

1, ..., μ. Then ξj (t, zj , 0) = zj , and the resolving functions are

αj (t, τ, zj , vj , 0) = αj (zj , vj )
26 Control of Moving Object Groups in a Conflict Situation

(vj , zj ) + (vj , zj )2 +  zj 2 (1 −  vj 2 )
= , j = 1, ..., μ.
 zj 2
Therefore, the pursuer’s control, which brings the jth system (1.31) to
zero, takes the form

u(τ ) = vj (τ ) − αj (zj (tj−1 ), vj (τ )) z(tj−1 ), τ ∈ [tj−1 , tj ), (1.33)

and the moments tj are defined by expression (1.32).

This control law (1.33) applies the strategy of parallel pursuit and is
closely related with the Apollonian circle [6], centered at the point, called
the Apollonius point. If μ = 1, then the Apollonian circle is the locus of the
capture points of the evader, under the condition the evader is moving with
maximum speed along a straight line, and the pursuer implements the strategy
of parallel pursuit (1.33).
As was shown in [6] in the case of ‘simple motions’ of the players and
μ = 1, in order to maximize the total time of capturing all the evaders, when
the pursuer sticks the strategy of parallel pursuit under fixed capture order,
each of the evaders should move with maximum speed along a straight line.
Let us call by the attainable circle of the jth evader the set

yj0 + βtj−1 ∂ S.

Each point of this circle and point x(tj−1 ) possesses an own Apollonius
circle with the corresponding center. Herewith, the centers of the Apollonius
circles, corresponding to the points of the attainable circle of the jth evader,
also form a circle (the circle of centers). Imagine that at each point of the
circle of centers the corresponding Apollonius circle is constructed. Then
the outer envelope of these circles is exactly the locus of the points of the
capture of player yj . The equation of this envelope in both polar and Cartesian
coordinates is presented in [6].
Hence, it is shown that if the order of captures is fixed, the pursuers
apply the law of parallel pursuit and the evaders are moving rectilinearly at
maximum speed, then the total capture time is functional depending only
on the constant controls of the evaders, and its maximization presents the
nonlinear finite-dimensional optimization problem [6].
To select the order of captures ensuring the least total time, the principle
of the shortest broken line was developed [6]. By the shortest broken line
is called a broken line of minimal length, starting from the initial position
of the pursuer and one at a time passing through all the initial positions
 References 27

of the evaders. In the case of ‘simple motions’, under certain assumptions,

the principle of the shortest broken line provides the smallest total capture
time [6].
For practical applications see [32].

1.7 Conclusion
An overview of contemporary research methods of the problem of decision
making and control in conflict conditions with the participation of moving
object groups is presented. The general problem of the game interaction of
groups with the help of decomposition is divided into separate problems of
group and successive pursuit with subsequent iterative repetition.
Methods for solving the mentioned problems, based on the method of
resolving functions, are presented. Their effectiveness is demonstrated, in
particular, on test examples from the monograph by R. Isaacs.
In so doing, the presence of phase coordinates for the state of objects is
taken into account, and the case of a non-fixed time of the game termination
is also highlighted. The situation of the environment by Pshenichnyi and the
principle of the shortest broken line make the results of group interaction
geometrically descriptive.

References
[1] A. A. Chikrii, ‘Escape Problem for Control Dynamic Objects’, Jour-
nal of Automation and Information Sciences, 1997, vol. 29, No. 6,
pp. 71–82.
[2] R. F. Isaacs, ‘Differential Games: Their Scope, Nature and Future’, J.
Opt. Theory Appl., 1969, No 3, p. 283–295.
[3] L. S. Pontryagin, ‘Linear Differential Games of Pursuit’, Matem.
Sbornik, 1980, 112, No 3, p. 307–330. (in Russian).
[4] B. N. Pschenichnyi, ‘Structure of Differential Games’, Dokl. Akad.
Nauk. SSSR, 1969, 184, No 2. (in Russian).
[5] N. N. Krasovskii, A. I. Subbotin, ‘Positional Differential Games’,
Nauka, Moscow, 1974 (in Russian).
[6] A. A. Chikrii, ‘Conflict controlled processes’, Naukova Dumka, Kiev,
1992 (in Russian).
[7] A. S. Locke, ‘Guidance’, D. Van Nostrand Company, Inc., Princeton,
1955.
28 Control of Moving Object Groups in a Conflict Situation

[8] B. S. Mordukhovich, ‘Approximation Methods in Problems of Opti-

mization and Control’, Nauka, Moscow, 1988 (in Russian).
[9] J. Warga, ‘Optimal Control of Differential and Functional Equations’,
Nauka, Moscow, 1977 (in Russian).
[10] B. N. Pschenitchnyi, ‘Simple pursuit by several objects’, Kibernetika,
1976, No. 3, pp. 145–146. (in Russian).
[11] N. L.Grigorenko, ‘Differential Games of Pursuit by Several Objects’
Izdat. Gos. Univ., Moscow, 1983. (in Russian).
[12] N. N. Petrov, N.N. Petrov, ‘On Differential Game “Kazaki -
Razboyniki”, Diff. Uravn., 1983, 19, No 8, p. 1366–1374 (in Russian).
[13] A. A. Chikrii, ‘Methods of Group Pursuit’, Stochastic Optimization, Int.
Conf., Edited by V.Arkin, A. Shiraev, R. Wets, Springer–Verlag, 1986,
pp. 632–640.
[14] A. A. Chikrii, S. F. Kalashnikova, L. A. Sobolenko, ‘A numerical
method for the solution of the successive pursuit problem’, Cybernetics,
1988, vol. 24, No. 1, pp. 53–59.
[15] B. N. Pshenichnyi, A. A. Chikrii, I. S. Rappoport, ‘An efficient method
of solving differential games with many pursuers’, Soviet Math. Dokl.,
1981, vol. 23, No. 1, pp. 104–109.
[16] G. Ts. Dzyubenko, B. N. Pschenichnyi, ‘Discrete differential Games
with information lag’, Cybernetics, 1972, vol. 8, No. 6, pp. 947–952.
[17] M. S. Nikolskiy, ‘Linear Differential Games of Pursuit with Information
Time Lag’, Soviet Math. Dokl., 1971, 197, No. 5, pp. 1018–1021. (in
Russian).
[18] L. A.Vlasenko, A. D. Myshkis, A. G. Rutkas, ‘On a class of differ-
ential equations of parabolic type with impulsive action’, Differential
Equations, 2008, 44, No. 2, pp. 231–244.
[19] A. A. Chikrii, ‘Game Dynamic Problems for Systems with Fractional
Derivatives’, Pareto-Optimality, Game Theory and Equilibria ,
2008, vol.17, New York, Springer, pp. 349–387.
[20] L. A. Vlasenko, A. G. Rutkas, ‘Stochastic impulse control of parabolic
systems of Sobolev type’, Differential Equations, 2011, 47, No. 10,
pp. 1498–1507.
[21] V. A. Pepelyaev, Al. A. Chikrii, K. A. Chikrii, ‘On nonstationary prob-
lems of motion control in conflict situation’, Journal of Automation and
Information Sciences, 2019, 51, No. 7, pp. 55–66.
[22] A. N. Khimich, K. A. Chikrii, ‘Game approach problems for dynamic
processes with impulse controls’, Cybernetics and Systems Analysis,
2009, 45, No 1, pp. 123–140.
 References 29

[23] I. Yu. Krivonos, Al. A. Chikrii, K. A. Chikrii, ‘Some modifications to

the pontryagin’s condition in nonstationary quasilinear games’, Journal
of Automation and Information Sciences, 2013, 45, No. 11, pp. 15–21.
[24] A. A. Chikrii, ’Linear problem of avoiding several pursuers’, Engineer-
ing Cybernetics, 1976, 14, No 4, pp. 38–42.
[25] A. A Chikrii, I. S. Rappoport, ‘Systems analysis method of resolving
functions in the theory of conflict-controlled processes’, Cybernetics
and Systems Analysis, 2012, 48, No 4, pp. 512–531.
[26] O. Hajek, ‘Pursuit games’, New York: Academic Press, 1975.
[27] M. S Nikolskiy, ‘The first direct L. S. Pontryagin method in differential
games’, Izdatelstvo MGU, Moscow, 1984.
[28] G. T. Chikrii, ‘On the problem of approach for damped oscillations’,
Journal of Automation and Information Sciences, 2009, 41, No. 10,
pp.1–9.
[29] A. G. Nakonechnyi, S. O. Mashchenko, V. K. Chikrii, ‘Motion con-
trol under conflict condition’, Journal of Automation and Information
Sciences, 2018, 50, No. 1, pp. 54–75.
[30] A. A. Chikrii, V. K. Chikrii, ‘Image Structure of Multivalued Map-
pings in Game Problems of Motion Control’, Problemi upravlenia i
informatiki, 2016, No 2, pp. 65–78.
[31] A. O. Chikrii, G. T. Chikrii, ‘Matrix Resolving Functions in Dynamic
Games of Approach’, Cybernetics and System Analysis, 2014, 50,
No. 2, pp. 201–217.
[32] Y. P. Kondratenko, V. M. Kuntsevich, A. A. Chikrii, V. F. Gubarev,
‘Advanced Control Systems: Theory and Applications’, River Publish-
ers, Series in Automation, Control and Robotics, 2021.
 2
Applications of Variational Analysis
to Controlled Sweeping Processes

Boris S. Mordukhovich

Department of Mathematics, Wayne State University, Detroit,

Michigan 48202, USA.
E-mail: [email protected].

Abstract
We present recent developments and applications of variational analysis and
generalized differentiation to rather new classes of optimal control problems
governed by discontinuous differential inclusions that arise in the so-called
sweeping processes and the like. Such problems are particularly important
for numerous applications to automatics control systems, engineering design,
hysteresis, etc. To study highly challenging control problems of this type,
we develop appropriate constructions of the method of discrete approxima-
tions, which allow us to investigate various theoretical and numerical aspects
of optimal control for discontinuous differential systems. In particular, we
establish new necessary optimality conditions for such problems including
appropriate extensions of the Euler–Lagrange conditions and the Pontryagin
maximum principle along with the optimality conditions of the novel type
specific for the problems under consideration. New applications to practical
engineering and mechanical models, robotics, traffic equilibria, etc. are also
discussed.

Keywords: Automatic control systems, optimal control, sweeping processes,

hysteresis, variational analysis, generalized differentiation, discrete approxi-
mations, optimality conditions, Euler–Lagrange and Hamiltonian inclusions,
Pontryagin’s maximum principle.

31
32 Applications of Variational Analysis to Controlled Sweeping Processes

2.1 Introduction and Discussions

Automatic control systems and related problems of control engineering
design are among the most important practical models that frequently appear
in applications. It has been well recognized that appropriate methods of opti-
mal control and differential games play a major role in theoretical investiga-
tions and solving of such problems; see, e.g., the books [13, 23, 28, 29, 31, 32,
42, 44] and the references therein. Among advanced techniques in the study
of challenging issues appearing in automatic control systems, we highlight
methods of modern variational analysis and generalized differentiation.
Variational analysis has been well recognized as an active and fruitful
area of mathematics sciences that, on one hand, concerns the study of
optimization-related problems and, on the other hand, applies optimization,
perturbation, and approximation ideas to the analysis of a broad range of
problems that may not be of a variational nature. This area can be considered
as an outgrowth of the classical calculus of variations, optimal control the-
ory, and mathematical programming, where the focus is on optimization of
functions relative to various constraints and on the solution sensitivity with
respect to perturbations.
A characteristic feature of modern variational analysis is a broad involve-
ment of objects with nonsmooth structures (nondifferentiable functions, sets
with nonsmooth boundaries, and set-valued mappings), which appear nat-
urally and frequently in the framework of optimization, equilibrium, and
control problems while being often generated by the usage of variational
principles and techniques. Nonlinear systems and variational principles in
applied sciences also give rise to nonsmooth structures and motivate the
developments of new forms of analysis that rely on generalized differenti-
ation. Starting with nonsmooth convex functions, subgradient mappings are
inevitably set-valued and thus require tools of set-valued analysis for their
study and applications.
The last two decades have witnessed rapid progress in developing and
applications of variational analysis and generalized differentiation. Starting
with the fundamental monograph by Rockafellar and Wets [45] devoted to
finite-dimensional aspects of variational analysis and their applications to
constrained optimization, an enormous amount of publications on the subject
appeared and great many applications were developed to various branches
of mathematical sciences as well as to engineering, economics, biology,
environmental and behavioral sciences, operations research, computer sci-
ences, robotics, etc. We refer the reader to [45] and the author’s monographs
 2.1 Introduction and Discussions 33

[36, 37], with the vast bibliographies and commentaries therein, for the gen-
esis of ideas, basic constructions and results, and applications of variational
analysis and generalized differentiation in finite and infinite dimensions. In
particular, the second volume of [36] contains a rather comprehensive account
of the methods and results of variational analysis applied to nonsmooth
controlled systems governed by ordinary differential, functional differential,
and partial differential equations and Lipschitzian differential inclusions with
a variety of applications.
New trends in variational analysis largely relate to the investigation and
solving of meaningful problems, which often come from applications. A
broad class of such problems are governed by discontinuous differential
inclusions introduced originally by Moreau [40, 41] in the form of the
sweeping process

ẋ(t) ∈ −N (x(t); C(t)) a.e. t ∈ [0, T ] with x(0) := x0 ∈ C(0), (2.1)

where ‘a.e.’ means as usual ‘almost everywhere’ in the sense of Lebesgue

measure on [0, T ], and where N (x; C) stands for the normal cone of convex
analysis defined by

N (x; C) := {v ∈ Rn | v, y − x ≤ 0, y ∈ C} if x ∈ C and

N (x; C) := ∅ if x ∈
/C (2.2)

for the continuously moving convex set C = C(t) at the point x = x(t). The
primal motivation of [41] mainly came from problems of elastoplasticity, but
over the years the sweeping process and its modifications have been devel-
oped in dynamical system theory with many applications to various areas of
mechanics, engineering, robotics, economics, traffic equilibria, etc.; see, e.g.,
the excellent recent survey [8] with the references therein. Independently,
basic theory and important engineering applications of sweeping processes
were largely developed (not under this name) in the Soviet literature by
Krasnosel’skii, Pokrovskii, and their followers for systems of hysteresis; see
the English book [30] with many references to the previous publications in
Russian.
Optimization and control problems for sweeping processes were formu-
lated much latter. A primal reason for this situation is that the Cauchy problem
for the sweeping process (1) has a unique solution, and so there is nothing to
optimize. To the best of our knowledge, optimal control problems for sweep-
ing processes were first defined with controls functions acting in additive
perturbations as in Edmond and Thibault [22]. However, the focus of [22], as
34 Applications of Variational Analysis to Controlled Sweeping Processes

well as of subsequent publication in this direction (see, e.g., Tolstonogov [48]

and related chapters), was on the existence of optimal controls and relaxation
procedures, while not on deriving necessary optimality conditions that have
always been at the core of the calculus of variation and optimal control.
Necessary optimality conditions for controlled sweeping processes were
first established less than ten years ago in Colombo et al. [16] (see also [17])
for a novel class of problems with control functions acting in the moving set
formalized as C(t) = C(u(t)) on [0, T ]. Another type of controlled sweeping
processes was introduced in Brokate and Krejčcí [9], with deriving necessary
optimality conditions, in the setting where control functions entered a linear
ODE system adjacent to the sweeping dynamics. First necessary optimality
conditions for sweeping control problems with control actions in dynamic
perturbations were derived in Cao and Mordukhovich [11]. After these initial
works, and especially in the very recent years, optimal control theory for
various types of controlled sweeping processes has been strongly developed
in many publications with establishing new necessary optimality conditions
such that some of them do not have any analogs in the classical theory of
optimal control; see, e.g., [2, 4, 10, 12, 15–20, 26, 52] among other chapters.
It is important to emphasize that all the aforementioned types of optimal
control problems dealing with the sweeping differential inclusions (2.1) are
highly different from optimal control systems governed by smooth ordinary
differential equations as in the classical monograph by Pontryagin et al.
[42] and by their nonsmooth counterparts, as well as by set-valued problems
governed by Lipschitzian differential inclusions of the type

ẋ(t) ∈ F (x(t)) a.e. t ∈ [0, T ], (2.3)

where F (x) is a Lipschitz continuous set-valued mapping (for simplicity we

consider here only autonomous systems); see, e.g., the books [14, 35, 36, 43,
46, 50] for more details and references. It is clear that the latter formalism
covers the ODE control systems

ẋ(t) = f (x(t), u(t)), u(t) ∈ U a.e. t ∈ [0, T ], (2.4)

where the control set U in (4) may depend also of the state U = U (x) and
thus reflects feedback laws. The sweeping inclusions (2.1) are dramatically
different from (2.3), and hence from (2.4), due to highly non-Lipschitzian
(actually discontinuous) structure of the sweeping dynamics (2.1), which
does not allow us to employ the conventional techniques developed earlier
in optimal control theory.
 2.1 Introduction and Discussions 35

Another, even more essential difference between (2.1) and the dynamics
in (2.3) and (2.4) is that in the case of controlled moving sets C(t) = C(u(t))
in (2.1), the right-hand site in (2.1) is not fixed as in (2.3) and in (2.4) with
F (x) := f (U (x)). Thus optimizing (2.1) is actually a problem of shape
optimization.
The main goal of this chapter is to discuss some recent results in dynamic
optimization of controlled sweeping processes and some of their applications.
We focus on the following two types of the sweeping dynamics: with controls
acting in moving sets and with controls entering additive perturbations.
These two classes of sweeping optimal controls and the necessary optimality
conditions obtained for them are essentially different from each other.
Nevertheless, we derive necessary optimality conditions for both of these
classes of sweeping control problems by using the method of discrete approx-
imations and advanced tools of generalized differentiation in variational
analysis. Recall that the discrete approximation approach to obtain optimality
conditions for continuous-time variational problems goes back to Euler in the
classical calculus of variations. This method was used by Pshenichnyi [43]
to optimize Lipschitzian differential inclusions governed by set-valued map-
pings with convex graphs and also with convex values under rather restrictive
assumptions. Those restrictions have been overcome for convex-valued Lip-
schitzian differential inclusions, with the usage of generalized differentiation
developed by the author (see Section 2), in the books by Mordukhovich
[35] and Smirnov [46]; see also the references therein. Then the author
[36] established, by employing again the method of discrete approximations,
refined optimality conditions for Lipschitzian differential inclusions without
any convexity assumptions. Below we present major results on necessary
optimality conditions obtained by appropriate developments of the method
of discrete approximations for the aforementioned classes of optimal con-
trol problems for sweeping differential inclusions of highly discontinuous
dynamics.
The rest of the chapter is organized as follows. In Section 2 we review
basic tools of first-order and second-order generalized differentiation in
variational analysis that is broadly used in the formulations and proofs of
the main necessary optimality conditions given below.
Section 3 addresses optimal control problems for sweeping processes
with control functions entering moving sets that generate the sweeping
dynamics. Here we establish necessary optimality conditions of generalized
Euler–Lagrange and Hamiltonian types, which yield a new version of the
Weierstrass–Pontryagin maximization condition in terms of the modified
36 Applications of Variational Analysis to Controlled Sweeping Processes

Hamiltonian while an expected form of the Pontryagin maximum principle

fails.
Section 4 deals with controlled sweeping processes, where control func-
tions appear in additive perturbations of the dynamics. A variety of necessary
optimality conditions obtained for such systems include, among other rela-
tionships of the new type, an appropriate version of the Pontryagin maximum
principle in the class of measurable controls. The concluding Section 5
presents selected applications of the achieved results and discuses some
perspectives of the ongoing and future research.
Throughout the chapter we use the standard notation of variational
analysis and optimal control; see, e.g., the books [36, 45, 50].

2.2 Generalized Differentiation of Variational Analysis

This section overviews the basic generalized differential constructions, which
play a central role of many aspects of variational analysis and its applications.
These constructions were mainly introduced by the author in different times
and then were largely developed by him and his collaborators along with
other researchers in this field. The reader can find more details, comments,
and references in the books [35–36, 37, 45].
Employing the geometric approach to generalized differentiation [35–37],
we start with our basic concept of the normal cone to a locally closed set
that induces the corresponding generalized differentiability notions for nons-
mooth functions and (single-valued and set-valued) mappings. Note that our
discrete approximation technique requires considering normals to nonconvex
sets even in the case of sweeping processes generated by convex moving sets
as in (2.1). The (basic, limiting, Mordukhovich) normal cone to a nonempty
locally closed set Ω ⊂ Rn at x̄ ∈ Rn is defined by
⎧
⎪
⎨{v ∈ R | ∃xk → x̄, αk ≥ 0, wk ∈ Π(xk ; Ω), αk (xk − wk )
n
⎪
N (x̄; Ω) := → v if x̄ ∈ Ω,
⎪
⎪
⎩∅ otherwise,
(2.5)
where Π(x; Ω) stands for the nonempty Euclidean projector of x onto Ω. If Ω
is convex, the normal cone (2.5) reduces to the normal cone of convex anal-
ysis (2.2), but in general the multifunction x⇒N (x; Ω) is nonconvex-valued
while satisfying a full calculus together with the associated subdifferential
of extended-real-valued functions and coderivative of set-valued mappings
 2.2 Generalized Differentiation of Variational Analysis 37

considered below. Such a calculus is due to variational/extremal principles

of variational analysis; see [36, 37, 45] for more details.
Given a set–valued mapping F : Rn ⇒ Rq and a point (x̄, ȳ) ∈ gph F
from its graph
gph F := {(x, y) ∈ Rn × Rq | y ∈ F (x))},
the coderivative D∗ F (x̄, ȳ) : Rq ⇒ Rn of F at (x̄, ȳ) is defined by
D∗ F (x̄, ȳ)(u) := {v ∈ Rn | (v, −u) ∈ N ((x̄, ȳ); gph F )}, u ∈ Rq ,
(2.6)
where ȳ is omitted in the notation when F : Rn → Rq is single-valued.
If F is C 1 -smooth around x̄, we have D∗ F (x̄)(v) = {∇F (x̄)∗ v} via the
adjoint Jacobian matrix. In general, the coderivative (2.6) is a positively
homogeneous set-valued mapping satisfying extended calculus rules and
providing fullcharacterizations of the fundamental well-posedness properties
in variational analysis concerning Lipschitzian stability, metric regularity, and
linear openness (covering); see [36, 45].
Given an extended-real-valued function φ : Rn → R := (−∞, ∞] finite
at x̄, i.e., with x̄ ∈ dom φ, the (first-order) subdifferential of φ at x̄ is
introduced geometrically by
∂φ(x̄) := {v ∈ Rn | (v, −1) ∈ N ((x̄, φ(x̄)); epi φ)} (2.7)
via the normal cone (2.5) to the epigraphical set epi φ := {(x, α) ∈
Rn+1 | α ≥ φ(x)}. If φ(x) := δΩ (x), the indicator function of a set Ω
equal to 0 for x ∈ Ω and to ∞ otherwise, we get ∂φ(x̄) = N (x̄; Ω).
Fixing v̄ ∈ ∂φ(x̄), the second-order subdifferential (or generalized Hessian)
∂ 2 φ(x̄, v̄) : Rn ⇒ Rn of φ at x̄ relative to v̄ is defined as the coderivative of
the first-order subdifferential by
∂ 2 φ(x̄, v̄)(u) := (D∗ ∂φ)(x̄, v̄)(u), u ∈ Rn , (2.8)
When φ is twice continuoisly differentiable around x̄, then (2.8) reduces to
the classical (symmetric) Hessian matrix
∂ 2 φ(x̄)(u) = {∇2 φ(x̄)u} f or all u ∈ Rn .
In what follows, we also use partial versions of the above subdifferential
constructions for functions of two variables φ : Rn × Rm → R. Define the
partial first-order subdifferential mapping for φ(x, w) with respect to x by
∂x φ(x, w) := { set of subgradients v ∈ Rn of φw := φ(·, w) at x}
= ∂φw (x)
38 Applications of Variational Analysis to Controlled Sweeping Processes

and then, for (x̄, w̄) ∈ dom φ and v̄ ∈ ∂x φ(x̄, w̄), introduce the partial
second-order subdifferential of φ with respect to x at (x̄, w̄) relative to v̄ by
∂x2 φ(x̄, w̄, v̄)(u) := (D∗ ∂x φ)(x̄, w̄, v̄)(u) f or all u ∈ Rn . (2.9)
If φ is twice continuously differentiable around (x̄, w̄), we have
∂ 2 φ(x̄, w̄)(u) = {(∇2xx φ(x̄, w̄)∗ u, ∇2xw φ(x̄, w̄)∗ u)} f or all u ∈ Rn .
Consider now the following class of parametric constraint systems
S(w) := {x ∈ Rn | ψ(x, w) ∈ Θ}, w ∈ Rm , (2.10)
generated by a vector function ψ : Rn × Rm → Rs and a set Θ ⊂ Rs , and
then associate with (2.10) the normal cone mapping N : Rn × Rm ⇒ Rn
defined by
N (x, w) := N (x; S(w)) f or x ∈ S(w). (2.11)
It is easy to see that the mapping N in (2.11) admits the representation
N (x, w) = ∂x φ(x, w) with φ(x, w) := (δΘ ◦ ψ)(x, w) (2.12)
via the composition of ψ and the indicator function δΘ of the set Θ. We get
from (2.12) and the second-order subdifferential construction (2.9) that
∂x2 φ(x̄, w̄, v̄)(u) = D∗ N (x̄, w̄, v̄)(u) f or any v̄ ∈ N (x̄, w̄) and u ∈ Rn .
The following second-order calculus result can be derived from [39,
Theorem 3.1] applied to the composition in (2.12). This computation
foprmula plays an important role in the our applications to establishing
optimality conditions for controlled sweeping processes.
Theorem 2.13 (second-order subdifferential chain rule). Let ψ be C 2 -
smooth around (x̄, w̄) with the partial Jacobian matrix ∇x ψ(x̄, w̄) of full
rank. Then for each v̄ ∈ N (x̄, w̄) there exists a unique vector p̄ ∈
NΘ (ψ(x̄, w̄)) := N (ψ(x̄, w̄); Θ) satisfying
∇x ψ(x̄, w̄)∗ p̄ = v̄
and such that the coderivative of the normal cone mapping is calculated for
all u ∈ Rn by
 
∇ 2 p̄, ψ(x̄, w̄)
D∗ N (x̄, w̄, v̄)(u) = u + ∇ψ(x̄, w̄)∗ D∗ NΘ (ψ(x̄, w̄), p̄)
xx
∇2xw p̄, ψ(x̄, w̄)
(∇x ψ(x̄, w̄)u).
 2.3 Dynamic Optimization via Controlled Moving Sets 39

2.3 Dynamic Optimization via Controlled Moving Sets

The results of this section are mostly based on the recent chapter [26], which
addresses a general class of sweeping processes with controlled moving sets
defined as inverse images of closed subsets of finite-dimensional spaces
under nonlinear differentiable mappings that depend on both state and control
variables. Such problems naturally arise in the study of and applications to
hysteresis and rate-independent systems.
The problem under consideration here is formulated as follows:
 T
minimize J[x, u] := ϕ(x(T )) + (t, x(t), u(t), ẋ(t), u̇(t))dt (2.14)
0

over absolutely continuous control actions u(·) and the corresponding abso-
lutely continuous trajectories x(·) of the sweeping differential inclusion
ẋ(t) ∈ f (t, x(t)) − N (g(x(t)); C(t, u(t))) a.e. t ∈ [0, T ],
x(0) = x0 ∈ C(0, u(0)), (2.15)
where the controlled moving set is given by
C(t, u) := {x ∈ Rn | ψ(t, x, u) ∈ Θ}, (t, u) ∈ [0, T ] × Rm , (2.16)
with f : [0, T ] × Rn → Rn , g : Rn → Rn , ψ : [0, T ] × Rn × Rm → Rs ,
and Θ ⊂ Rs . The sets C(t, u) may not be convex, and the normal cone in
(2.15) is understood in the sense of (2.5). It follows from (2.15) due to the
normal cone definition that the formulated optimal control problem contains
the pointwise constraints on both state and control functions given by
ψ(t, g(x(t)), u(t)) ∈ Θ f or all t ∈ [0, T ].
Such constraints have been realized among the most challenging ones even in
standard optimal control theory for smooth ordinary differential equations.
Let us now formulate ourstanding assumptions in this section:
(H1) There is Lf > 0 such that f (t, x) − f (t, y) ≤ Lf x − y for all
x, y ∈ Rn , t ∈ [0, T ] and the mapping t → f (t, x) is a.e. continuous
on [0, T ] for each vectorx ∈ Rn .
(H2) There isLg > 0 such that g(x)−g(y) ≤ Lg x−y for all x, y ∈ Rn .
(H3) For each (t, u) ∈ [0, T ] × Rm , the mapping ψt,u (x) := ψ(t, x, u) is
twice continuously differentiable around the reference points with the
surjective derivative ∇ψt,u (x) satisfying
∇ψt,u (x) − ∇ψt,v (x) ≤ Lψ u − v
40 Applications of Variational Analysis to Controlled Sweeping Processes

with the uniform Lipschitz constant Lψ . Also we assume that the

mapping t → ψ(t, x) is a.e. continuous on [0, T ] for each x ∈ Rn
and u ∈ Rm .
(H4) There are a number τ > 0 and a mapping ϑ : Rn × Rn × Rn × Rm →
Rm locally Lipschitz continuous and uniformly bounded on bounded
−1
sets such that for all t ∈ [0, T ], v̄ ∈ N (ψ(t,ū) (x̄); Θ), and x ∈ ψ(t,u) (Θ)
with u := ū + ϑ(x − x̄, x, x̄, ū) there exists v ∈ N (ψ(t,u) (x); Θ)
satisfying v − v̄ ≤ τ x − x̄.
(H5) The objective functions ϕ : Rn → R := [−∞, ∞) and (t, ·) :
R2(n+m) → R in (2.14) are bounded from below and lower semicon-
tinuous around a given feasible solution to (P ) for a.e. t ∈ [0, T ], while
the running costs  is a.e. continuous in t and is uniformly majorized
by a summable function on [0, T ].
(H6) The set Θ in (2.16) has nonempty interior and locally closed around
the reference points.

Assumption (H4) is technical being the most restrictive. As shown in [26], it

is automatically satisfied in the case of polyhedral moving sets

C(t) := {x ∈ Rn | ui (t), x ≤ wi (t), i = 1, . . . , m}

controlled by both ui (t) and wi (t) as in [17], and also in various nonconvex
settings.
Our approach to the study of the sweeping control problem formulated in
(2.14)–(2.16) is based on the method of discrete approximations containing
the following major steps:
Step I: Construct a well-defined family of discrete approximations involving
a finite-difference replacement of the derivative ẋ in (2.15). The key point of
this step is to verify the possibility to approximate any feasible trajectory
of (2.15) by feasible trajectories of discrete-time systems in a topology
yielding the a.e. convergence of the extended discrete derivatives alonmg a
subsequence. We address not only qualitative aspects of well-posedness but
also quantitative ones with estimating error bounds, convergence rates, etc.,
which are of undoubted numerical values. Achieving these goals would lead
us then to the strong W 1,2 -norm approximation of a given local minimizer for
the continuous-time problem under consideration by a sequence of optimal
solutions to the discrete-time problems that are piecewise linearly extended
to the whole interval [0, T ].
 2.3 Dynamic Optimization via Controlled Moving Sets 41

Step II: For each fixed step of discretization, the approximating discrete-time
problems can be reduced to nondynamic problems of mathematical pro-
gramming with various constraints, including increasingly many geometric
constraints. The latter problems are finite-dimensional of increasing dimen-
sions. Powerful tools of generalized differentiation in variational analysis
overviewed in Section 2 can be employed for deriving discrete-time neces-
sary optimality conditions in the approximation problems, even without any
Lipschitzian and convexity assumptions, by using appropriate calculus rules.
However, dealing with the graphical structure of the geometric constraints
in the approximation problems requires robust generalized differential con-
structions enjoying full calculus that should be subtle and small enough to be
efficiently applied to graphical sets. Note to this end, note that applying the
convexified normal cone and related constructions by Clarke [14] to graphs of
mappings often gives us the whole space or its subspace of maximal dimen-
sion; see [36, 45]. On the other hand, the nonconvex generalized differential
constructions from Section 2 satisfy all the required properties and can be
successfully used.
Step III: The concluding step in deriving necessary conditions for optimal
solutions of (2.14)–(2.16) is the passage to the limit from those for discrete-
time problems obtained in Step II with employing the strong convergence
of discrete optimal solutions obtained in Step I. This final step is highly
involved, since it requires justifying an appropriate convergence of dual arcs
as trajectories of adjoint differential inclusions. In this way we arrive at
the necessary optimality conditions for sweeping control problems presented
below.
Let us now formulate the notion of local minimizers for which we derive
necessary optimality conditions in what follows.
Definition 2.17 (local minimizers for controlled sweeping processes).
Let the pair (x̄(·), ū(·)) be feasible to problem (2.14)–(2.16) under the
imposed standing assumptions. We say that (x̄(·), ū(·)) be a local W 1,2 ×
W 1,2 -minimizer for this problem if x̄(·) ∈ W 1,2 ([0, T ]; Rn ), ū(·) ∈
W 1,2 ([0, T ]; Rm ), and
J[x̄, ū] ≤ J[x, u] f or all x(·) ∈ W 1,2 ([0, T ]; Rn ) and
u(·) ∈ W 1,2 ([0, T ]; Rm )
sufficiently close to (x̄(·), ū(·)) in the corresponding W 1,2 -norm topology.
Note that this notion, while being rather specific for the class of control
problems under consideration, occupies an intermediate position between
42 Applications of Variational Analysis to Controlled Sweeping Processes

the conventional notions of weak and strong local minimizers in variational

problems; see [36, Chapter 6] for more discussions in the framework of
Lipschitzian differential inclusions. Observe also the choice of the class of
feasible solutions to problem (2.14)–(2.16) is due to the known existence
theorems for the sweeping dynamics; compare, e.g., [12, 33, 48].
As has been well understood in the calculus of variations and optimal
control, starting with the pioneering chapters by Bogolyubov [6] and Young
[51], finishing limiting procedures in variational arguments require some
relaxation stability under the velocity convexification, which in fact holds
automatically in fairly general settings; see [14, 36, 47, 50] for Lipschitzian
differential inclusions and [22, 48] for non–Lipschitzian frameworks of
sweeping processes.
Following this idea, let us define the mapping

F = F (t, x, u) := f (t, x) − N (g(x); C(t, u)) (2.18)

and formulate the relaxed optimal control problem for (2.14)–(2.16) as

follows:
 T
minimize J[x, u] := ϕ(x(T )) + F (t, x(t), u(t), ẋ(t), u̇(t))dt, (2.19)
0

where (t, x, u, ·, ·) is defined as the largest l.s.c. convex function majorized

by (t, x, u, ·, ·) on the convex closure of the set F in (2.18) with  :=
∞ otherwise. Then we say that the pair (x̄(·), ū(·)) is a relaxed local
W 1,2 × W 1,2 -minimizer for (2.14)–(2.16) if in additions to the conditions
of Definition 2.17 we have J[x̄, ū) = J[x̄, ū]. Observe that, contrary to the
original sweeping control system in (2.14)–(2.16), the convexified structure
of the relaxed dynamic optimization problem (2.19) allows us to establish the
existence of global optimal solutions in the prescribed classes of controls and
trajectories; cf. [12, 33, 48].
As we see, there is no difference between the original and relaxed
problems if the normal cone in (2.18) is convex and the integrand  in
(2.14) is convex with respect to velocity variables. On the other hand, the
nonatonomicity of Lebesgue’s measure on [0, T ] and the differential inclusion
structure of the sweeping process (2.15) are instrumental to conclude, due to
the classical Lyapunov convexity theorem [34] (see also [5, 27]) that in many
important situations local minimizer of the types under consideration is also
a relaxed one. Without delving into details here, we just mention that the
possibility to derive such a local relaxation stability from [48, Theorem 4.2]
 2.3 Dynamic Optimization via Controlled Moving Sets 43

for strong local minimizers of problem (2.14)–(2.16), provided that the

controlled moving sets C(t, u) in (2.16) are convex and continuous.
Now we present our first major result establishing generalizesd the
Euler–Lagrange formalism for the class of sweeping control problems
(2.14)–(2.16). The reader can see that, besides extended conditions of the
Euler–Lagrange and transversality types, the obtained theorem contains the
necessary optimality conditions of novel types, which are specific for the
controlled sweeping dynamics. The proof of this theorem, which follows the
lines in the proof of [26, Theorem 4.3], realizes the steps (I)–(III) of the
discrete approximation method and machinery of variational analysis with
the particular usage of the second-order calculations from Theorem 2.13.
For simplicity the theorem is formulated in the case of (2.14)–(2.16) where
g(x) := x, f := 0 while ψ and  do not depend on t. Recall the the symbol
‘co’ stands for the convex hull of the set in question.
Theorem 3.2 (necessary conditions for controls in moving sets). Let
(x̄(·), ū(·)) be a relaxed local W 1,2 × W 1,2 -minimizer local minimizer for
problem (2.14)–(2.16). In addition to the standing assumptions, suppose that
ψ = ψ(x, u) is C 2 -smooth with respect to both variables while the cost func-
tions ϕ and  are locally Lipschitzian around the corresponding components
of the optimal solution. Then there exist a multiplier λ ≥ 0, an adjoint
arc p(·) = (px , pu ) ∈ W 1,2 ([0, T ]; Rn × Rm ), a signed vector measure
γ ∈ C ∗ ([0, T ]; Rs ), as well as pairs (wx (·), wu (·)) ∈ L2 ([0, T ]; Rn × Rm )
and (v x (·), v u (·)) ∈ L∞ ([0, T ]; Rn × Rm ) with
(wx (t), wu (t), v x (t), v u (t)) ∈ co∂(x̄(t), ū(t), x̄(t),
˙ ˙
ū(t)) a.e. t ∈ [0, T ]
satisfying the following necessary optimality conditions:

• PRIMAL–DUAL DYNAMIC RELATIONSHIPS:

 
∇2xx η(t), ψ(x̄(t), ū(t))
ṗ(t) = λw(t) +
∇2xw η(t), ψ(x̄(t), ū(t))
(−λv x (t) + q x (t)) a.e. t ∈ [0, T ], q u (t) = λv u (t) a.e. t ∈ [0, T ],
where η(·) ∈ L2 ([0, T ]; Rs ) is a uniquely determined function from
˙
x̄(t) = −∇x ψ(x̄(t), ū(t))∗ η(t) a.e. t ∈ [0, T ]
with η(t) ∈ N (ψ(x̄(t), ū(t)); Θ), and where q : [0, T ] → Rn × Rm
is a function of bounded variation on [0, T ] with its left-continuous rep-
resentative given, for all t ∈ [0, T ] except at most a countable subset,
44 Applications of Variational Analysis to Controlled Sweeping Processes

by 
q(t) = p(t) − ∇ψ(x̄(τ ), ū(τ ))∗ dγ(τ ).
[t,T ]
• MEASURED CODERIVATIVE CONDITION: Considering the
] t-dependent outer limit

γ(B)
Limsup (t) := y ∈ Rs | ∃ sequence Bk ⊂ [0, 1] with
|B|→0 |B|

γ(Bk )
t ∈ IBk , |Bk | → 0, →y
|Bk |
over Borel subsets B ⊂ [0, 1] with the Lebesgue measure |B|, for a.e.
t ∈ [0, T ] we have

D∗ NΘ (ψ(x̄(t), ū(t)), η(t))(∇x ψ(x̄(t), ū(t))(q x (t) − λv x (t)))∩

γ(B)
Limsup (t) = ∅.
|B|→0 |B|

• TRANSVERSALITY CONDITION AT THE RIGHT ENDPOINT:

−(px (T ), pu (T )) ∈ λ(∂ϕ(x̄(T )), 0)+∇ψ(x̄(T ), ū(T ))NΘ ((x̄(T ), ū(T )).

• MEASURE NONATOMICITY CONDITION: Whenevert ∈ [0, T )

withψ(x̄(t), ū(t)) ∈ intΘ there is a neighborhood Vt of t in [0, T ] such
that γ(V ) = 0 for any Borel subset V of Vt .

Nontriviality condition:

λ + sup p(t) + γ = 0 with γ := sup x(s)dγ.
t∈[0,T ] xC([0,T ] =1 [0,T ]

Suppose in addition that η(T ) is well defined and that θ = 0 is the only vector
for which

θ ∈ D∗ NΘ (ψ(x̄(T ), ū(T )), η(T ))(0),

∇ψ(x̄(T ), ū(T ))∗ θ ∈ ∇ψ(x̄(T ), ū(T ))NΘ (x̄(T ), ū(T )).

Then the above necessary optimality conditions hold with the stronger
nontriviality

λ + mes{t ∈ [0, T ]| q(t) = 0} + q(0) + q(T ) > 0.

 2.3 Dynamic Optimization via Controlled Moving Sets 45

Looking at the necessary optimality conditions obtained in Theorem 3.2,

we do not see there any maximization condition of the Weierstrass–
Pontryagin type, which is expected for continuous-time variational problems.
However, the new class of the sweeping control problems (2.14)–(2.16)
with controlled moving sets is largely different from conventional classes of
problems in dynamic optimization. Thus the reader should not be surprised by
the failure of the expected extension of the Pontryagin maximum principle for
problems of this type. We’ll construct below a (counter)example illustrating
this phenomenon, but first present necessary optimality conditions of the
novel Hamiltonian type that contain a modified (not expected a priori) form
of the maximum principle.
To proceed, consider the constraint set Θ in (2.16) given by

Θ = h−1 (Rl− ) := {z ∈ Rs | h(z) ∈ Rl− }, (2.21)

where h : Rs → Rl is a C 2 -smooth mapping. Define the collection of active

indexes
I(x, u) := {i ∈ {1, . . . , s}| ψi (x, u) = 0}.
It follows from the standing assumption (H3) that for each v ∈ −N (x; C(u))
there
 exists a unique collection {αi }i∈I(x,u) with αi ≤ 0 and v =
i∈I(x,u) αi [∇x ψ(x, u)]i . Given ν ∈ R , define the vector [ν, v] ∈ R by
s n

!
[ν, v] := νi αi [∇x ψ(x, u)]i
i∈I(x,u)

and introduce the modified Hamiltonian function

Hν (x, u, p) := sup{[ν, v], p| v ∈ −N (x; C(u))},

(x, u, p) ∈ Rn × Rm × Rn . (2.22)

Observe that in the linear case h(z) := Az −b in (2.21), the full rank assump-
tion in (H3) corresponds to the classical linear independence constraint
qualification (LICQ).
The following major results can be derived from Theorem 2.20 by using
the precise computation of the second-order construction D∗ NRs− .
Theorem 3.3 (novel Hamiltonian formalism and maximum principle for
optimal control in moving sets ). Consider the sweeping control problem
(2.14)–(2.16) in the frameworks of Theorem 2.20 with the set Θ given
by (2.21), where h : Rs → Rl is C 2 -smooth around the relaxed local
46 Applications of Variational Analysis to Controlled Sweeping Processes

W 1,2 × W 1,2 -minimizer z̄(t) := (x̄(t), ū(t)) for all t ∈ [0, T ]. Suppose
that the matrix ∇h(z̄(t)) is of full rank for all t ∈ [0, T ]. Then, in addition
to the necessary optimality conditions of Theorem 2.20, the new maximum
condition

[ν(t), x̄(t)],
˙ q x (t) − λv x (t) = Hν(t) (x̄(t), ū(t), q x (t) − λv x (t)) = 0 a.e.
t ∈ [0, T ] (20)

holds in terms of the modified Hamiltonian (2.22), where ν : [0, T ] → Rs is

taken from

ν(t) ∈ D∗ NRs− (h(ψ(x̄(t), ū(t))), μ(t))(∇x ψ(x̄(t), ū(t))(q x (t) − λv x (t)))∩

γ(B)
Limsup (t)
|B|→0 |B|

for a.e. t ∈ [0, T ] with a measurable vector function μ : [0, T ] → Rl

satisfying

μ(t) ∈ NRl (h(ψ(x̄(t), ū(t))) and

−

η(t) = ∇h(ψ(x̄(t), ū(t))∗ μ(t) a.e. t ∈ [0, T ].

Discussions on the Maximum Principle. Observe that our form of the new
Hamiltonian (2.22) is different from the conventional Hamiltonian function

H(x, p) := sup{p, v| v ∈ F (x)} (2.25)

used in optimal control of Lipschitzian differential inclusions ẋ ∈ F (x) that

extends the usual Hamiltonian functions in classical mechanics and optimal
control. We show below in Example 2.26 that the maximum principle via the
standard Hamiltonian (2.25) fails for the sweeping control problem in (2.14)–
(2.16). The main reason for this is that (2.25) does not reflects implicit state
constraints, which do not appear for Lipschitzian problems while being an
essential part of the sweeping dynamics.
Observe also that the new maximum principle form (2.24) incorporates
vector measures appearing through the measured coderivative condition of
Theorem 2.20. The fact that measures naturally arise in descriptions of neces-
sary optimality conditions in optimal control problems with state constraints
has been first realized by Dubovitskii and Milyutin [21] and since that has
been largely employed in optimal control. There are interesting connections
 2.3 Dynamic Optimization via Controlled Moving Sets 47

between our form of the maximum principle for controlled sweeping pro-
cesses and the Hamiltonian formalism in models of contact and nonsmooth
mechanics; see, e.g., [7].
Here is the promised example, which shows the failure of the conven-
tional maximum principle for sweeping optimal processes with controlled
moving sets and also illustrates the fulfillment of the new one obtained in
Theorem 2.23.
Example 3.4 (failure of the maximum principle for sweeping processes
with controlled moving sets). Consider problem (2.14)–(2.16) with u =
(a, b) ∈ R6 as a = (a1 , a2 ) ∈ R4 and b = (b1 , b2 ) ∈ R2 , x = (x1 , x2 ) ∈ R2 ,
x0 = (1, 1), T = 1,
x2 1
ϕ(x) = and (t, x, u, ẋ, u̇) := (ḃ21 + ḃ22 ).
2 2
The moving sets C(t, u) are defined by
C(t, u) := {x ∈ R2 | ai (t), x ≤ bi (t), i = 1, 2}.
Fix the a-controls as ā1 ≡ (1, 0) and ā2 ≡ (0, 1), and then deduce from
Theorem 2.20 the following relationships on [0, 1] involving only the b-
components of controls together with the corresponding state and adjoint
functions:
1. w(·) = 0, v x (·) = 0, v b (·) = (ḃ1 (·), ḃ2 (·))
 0 ⇒ qix (t) = 0, i = 1, 2
2. x̄˙ i (t) =
3. pb (·) is constant with nonnegative components, and −pxi (·) = λx̄(1) +
pbi (·)āi are also constant for both i = 1, 2.
˙ = pb + γ([t, 1]) for a.e. t ∈ [0, 1].
4. q x (t) = px − γ([t, 1]), q b (t) = λb̄(t)
5. λ + q(0) + p(1) = 0 with λ ≥ 0.
Observe first that the pair x̄(t) = (1, 1) and b̄(t) = 0 on [0, 1] satisfies the
necessary conditions with px1 = px2 = −1, pb1 = pb2 = γ1 = γ2 = 0, and
λ = 1. The conventional Hamiltonian function (2.25) reads now as
H(x, b, p) = sup{p, v| v ∈ −N (x; C((1, 0), (0, 1)), b)},
and we get by the direct calculation that
H(x̄(t), b̄(t), q x (t) − λv x (t))
= H((1, 1), (1, 1), (−1, −1))
= sup{(−1, −1), v| v ∈ −N ((1, 1); C(((1, 0), (0, 1)), (1, 1))}
= sup{(−1, −1), v| v1 ≤ 0, v2 ≤ 0} = ∞,
48 Applications of Variational Analysis to Controlled Sweeping Processes

while x̄(t),
˙ q x (t) − λv x (t) = 0. Thus the conventional maximum principle
fails in this example. At the same time, the new maximum condition (2.24)
from Theorem 2.23 holds trivially with the choice of ν(t) = 0 for all t ∈ [0, 1]
therein.

2.4 Sweeping Processes with Controlled Dynamics

In this section we consider a different type of controlled sweeping pro-
cesses, where measurable control functions enter dynamical perturbations.
Such problems can be viewed as far-going extensions of the standard class
of ODE optimal control and in fact reduce to the latter in the trivial (for
sweeping processes) case where the moving sets are absent. Employing
again the method of discrete approximations together with advanced tools of
variational analysis, we derive necessary optimality conditions for sweeping
control problems considered here that contain appropriate extensions of the
Pontryagin maximum principle along novel conditions of the types discussed
in Section 3 for problems with controlled moving sets. The material of the
section is mainly based on the recent chapter [19], where the reader can find
more details.
Here we consider the problem:
minimize J[x, u] := ϕ(x(T )) (2.27)
over pairs (x(·), u(·)) of measurable controls u(t) and absolutely continuous
trajectories x(t) on the fixed time interval [0, T ] satisfying the controlled
sweeping differential inclusion
ẋ(t) ∈ −N (x(t); C)+g(x(t), u(t)) a.e. t ∈ [0, T ], x(0) := x0 ∈ C ⊂ Rn ,
(2.28)
subject to the pointwise/hard control constraints given by
u(t) ∈ U ⊂ Rd a.e. t ∈ [0, T ]. (2.29)
The set C in (2.28) is a convex polyhedron given by

m
C := Ci with Ci := {x ∈ Rn | x∗i , x ≤ ci }. (2.30)
i=1

Observe that the sweeping differential inclusion (2.28) and the polyhedral
description (2.30) automatically yield the pointwise state constraints
x∗i , x(t) ≤ ci f or all t ∈ [0, T ].
 2.4 Sweeping Processes with Controlled Dynamics 49

It is said that a feasible pair (x̄(·), ū(·)) for (2.27)–(2.30) is a W 1,2 × L2 -

local minimizer for this problem if there exists a positive number ε such that
J[x̄, ū] ≤ J[x, u] for all feasible pairs (x(·), u(·)) satisfying the localization
condition
 T 
2
ẋ(t) − x̄(t)
˙ + u(t) − ū(t)2 dt < ε.
0

Our standing assumptions in this section are as follows:

(A1) The set U is compact and convex in Rd , while g(x, U ) is convex in Rn .
(A2) The cost function ϕ : Rn → R in (2.27) is C 1 -smooth around x̄(T ).
(A3) The perturbation mapping g : Rn × Rd → Rn in (2.28) is C 1 -smooth
around (x̄(·), ū(·)) and satisfies the sublinear growth condition

g(x, u) ≤ β(1 + x) f or all u ∈ U with some β > 0.

(A4) The vertices x∗i of (2.30) satisfy the linear independence constraint
qualification
!
[ αi x∗i = 0, αi ∈ R] ⇒ [αi = 0 f or all i ∈ I(x̄)}
i∈I(x̄)

along the trajectory x̄ = x̄(t) as t ∈ [0, T ], where I(x̄) := {i ∈

{1, . . . , m} | x∗i , x̄ = ci }.
Note that the usage of the relaxation procedure similar the one described in
Section 3 allows us to significantly relax the convexity assumptions in (A1).
Furnishing all the steps (I)–(III) of the method of discrete approximations
as discussed in Section 3 and using the generalized differential constructions
reviewed in Section 2, we arrive the following set of necessary optimality
conditions for the sweeping control problem (2.27)–(2.30). Note that an
important role in this device is played by explicit coderivative calculation
of the mapping on the right-hand side of (2.28) obtained in [25].
Theorem 2.31 (necessary conditions for sweeping processes with con-
trolled dynamics). Let (x̄(·), ū(·)) be a W 1,2 × L2 -local minimizer for
problem (2.27)–(2.30) under the assumptions in (A1)–(A4), where ū(·) is
of bounded variation (BV) with a right continuous representative on [0, T ].
Then there exist a multiplier λ ≥ 0, a measure γ = (γ1 , . . . , γn ) ∈
C ∗ ([0, T ]; Rn ) as well as adjoint arcs p(·) ∈ W 1,2 ([0, T ]; Rn ) and q(·) ∈
BV ([0, T ]; Rn ) such that λ + q(t)L∞ + p(T ) > 0 and the following
conditions are satisfied:
50 Applications of Variational Analysis to Controlled Sweeping Processes

• PRIMAL VELOCITY REPRESENTATION:

!
m
− x̄(t)
˙ = ηi (t)x∗i − g(x̄(t), ū(t)) f or a.e. t ∈ [0, T ], (2.32)
i=1

where η i (·) ∈ L2 ([0, T ]; R+ ) being uniquely determined by (2.32) and well

defined at t = T .
• ADJOINT SYSTEM:
ṗ(t) = −∇x g(x̄(t), ū(t))∗ q(t) f or a.e. t ∈ [0, T ],
where the dual arcs q(·) and p(·) are connected by the equation

q(t) = p(t) − dγ(τ )
(t,T ]

that holds for all t ∈ [0, T ] except at most a countable subset.

• MAXIMIZATION CONDITION:
ψ(t), ū(t) = max{ψ(t), u| u ∈ U } with ψ(t) := ∇u g(x̄(t), ū(t))∗
q(t) f or a.e. t ∈ [0, T ].

• COMPLEMENTARITY CONDITIONS:
x∗i , x̄(t) < ci ⇒ ηi (t) = 0 and ηi (t) > 0 ⇒ x∗i , q(t) = ci
for a.e. t ∈ [0, T ] including t = T and for all i = 1, . . . , m.
• RIGHT ENDPOINT TRANSVERSALITY CONDITIONS:
!
−p(T ) = λ∇ϕ(x̄(T )) + ηi (T )x∗i with
i∈I(x̄(T ))
!
ηi (T )x∗i ∈ N (x̄(T ); C).
i∈I(x̄(T ))

• MEASURE NONATOMICITY CONDITION: If t ∈ [0, T ) and

x∗i , x̄(t) < ci for all i = 1, . . . , m, then there is a neighborhood Vt of t
in [0, T ] such that γ(V ) = 0 for all the Borel subsets V of Vt .
Observe that the maximization condition of Theorem 2.31 provided an appro-
priate linearized version of the maximum principle for the class of optimal
control problems under consideration satisfying the imposed assumptions.
 2.5 Some Applications 51

2.5 Some Applications

In this section we briefly review selected applications of the obtained results
and also discuss some potential avenues for future developments.
Elastoplasticity and Systems of Hysteresis. Consider the model of this type
discussed in [3], but without any control actions. Let Q be a closed convex
subset of the 12 n(n + 1)-dimensional space of symmetric n × n tensors with
nonempty interior. Using the notation of [3], define the strain tensor = {}i,j
by := e + p , where e is the elastic strain and p is the plastic strain. The elastic
strain e depends on the stress tensor σ = {σ}i,j linearly, i.e., e = A2 σ, where
A is a constant symmetric positive-definite matrix. The principle of maximal
dissipation says that

˙p (t), z ≤ ˙p (t), σ(t) f or all z ∈ Q. (2.33)

As follows from [3], the variational inequality (2.33) is equivalent to the

sweeping process

ζ̇(t) ∈ −N (ζ(t); C(t)), ζ(0) = Aσ(0) − A−1 (0) ∈ C(0),

where ζ(t) := Aσ(t) − A−1 (t) and C(t) := −A−1 (t) + AQ. This can
be rewritten in the frame of the sweeping optimal control problem with
controlled moving sets considered in Section 3 where x := ζ, u := ,
ψ(x, u) := x + A−1 u, and Θ := AQ. Thus we can apply Theorems 2.20
and 2.31 to this class of hysteresis operators for the general elasticity domain
Q. In this way we cover various elastoplasticity models including those with
the Drucker-Prager, Mohr-Coulomb, Tresca, and von Mises yield criteria; see
[1]. More details with precise computations, examples, and discussions can
be found in [26].
Optimal Control of Sweeping Processes in Somew Models of Robotics.
Among important dynamical models arising in robotics, the mobile robot
model with obstacles has been well recognized and investigated. It is shown
in [24] that the dynamics of this model can be adequately described as a
sweeping process. An optimal control version of the mobile robot model with
obstacles has been recently developed in [18] as a sweeping control problem
with controlled dynamics studied in Section 4. This model concerns n mobile
robots (n ≥ 2) identified with safety disks in the plane of the same radius R.
The goal of each robot is to reach the target by the shortest path during a fixed
time interval [0, T ] while avoiding the other n − 1 robots that are treated by
this robot as obstacles.
52 Applications of Variational Analysis to Controlled Sweeping Processes

To formalize the model, consider the configuration vector x =

(x1 , . . . , xn ) ∈ R2n , where xi ∈ R2 is the center of the safety disk i with
coordinates (xi  cos θi , xi  sin θi ). This means that the trajectory xi (t) of
the i-robot/obstacle admits the representation

x̄i (t) = (x̄i (t) cos θi (t), x̄i (t) sin θi (t)) f or i = 1, . . . , n,

where the angle θi signifies the corresponding direction. According to the

model dynamics, at the moment of contacting the obstacle (one or more) the
robot in question keeps its velocity and pushes the other robots in contact to
go to the target with the same velocity and then to maintain their constant
velocities until reaching either other obstacles or the end of the process at
the final time t = T . In this framework, the constant direction θi of xi is the
smallest positive angle in standard position formed by the positive x-axis and
Oxi with taking the origin as the target point.
The optimal control problem corresponding to the described model is
formulated as:
1
minimize J[x, u] := x(T )2 ,
2
over the controlled sweeping dynamics

−ẋ(t) ∈ N (x(t); C) − g(x(t), u(t)) f or a.e. t ∈ [0, T ],

x(0) = x0 ∈ C, u(t) ∈ U a.e. on [0, T ],

where the controlled additive perturbations of the dynamics are modeled by

g(x(t), u(t)) = (s1 u1 (t) cos θ1 (t), s1 u1 (t) sin θ1 (t), . . . , sn un (t) cos θn (t),

sn un (t) sin θn (t)),

where the generating polyhedral set C ⊂ Rn is given by

C := {x ∈ Rn | xj∗ , x ≤ cj , j = 1, . . . , n − 1} with

xj∗ := ej − ej+1 , cj = −2R

for j = 1, . . . , n − 1 with (e1 , . . . , en ) being the orths in Rn , and where the
control set U is specified in [18] in particular settings.
Applying the necessary optimality conditions of Theorem 2.31 to this
problem leads to constructive algorithms of solving of the mobile mobile
model that we completely implemented in [18] to find optimal solutions in
 2.5 Some Applications 53

some interesting cases, while more serious practical applications require more
elaborations.
Traffic Equilibria and Crowd Motion Models. There are several practically
important models of traffic equilibria and related areas, where the dynamics
can be formulated as a sweeping process. One of the most interesting classes
of models of this type is known under the name of crowd motion models. In
what follows we consider some control versions of such problems, which can
be efficiently investigated and partly solved by using the necessary optimality
conditions discussed above.
The original motivation for crowd motion models came the modeling of
local interactions between participants in order to describe the dynamics of
pedestrian traffic. However, by now such models have been successfully used
to investigate more general classes of problems in engineering, operations
research, economics, etc.
The microscopic form of the crowd motion model is based on the follow-
ing two postulates. On one hand, each individual has a spontaneous velocity
that he/she intends to implement in the absence of other participants. On
the other hand, the actual velocity must be taken into account. The latter
one is incorporated via a projection of the spontaneous velocity into the
set of admissible/feasible velocities, i.e., those which do not violate certain
nonoverlapping constraints. A mathematical description of the uncontrolled
microscopic crowd motion model is given [49] in the sweeping process form
with the subsequent usage therein for numerical simulations and various
applications.
In the corridor version of the crowd motion model, the crucial nonover-
lapping condition is written in the form

Q0 = {x = (x1 , . . . , xn ) ∈ Rn , xi+1 − xi ≥ 2R},

where n ≥ 2 indicates the number of participants identified with rigid disks

of the same radius R in a corridor. The actual velocity field is described via
the projection operator by

ẋ(t) = Π(U (x); Cx ) f or a.e. t ∈ [0, T ], x(0) = x0 ∈ Q0 ,

with spontaneous velocities U (x) = (U0 (x1 ), . . . , U0 (xn )), x ∈ Q0 ,

satisfying

U (x) ∈ Nx + ẋ(t) f or a.e. t ∈ [0, T ], x(0) = x0 ,

54 Applications of Variational Analysis to Controlled Sweeping Processes

where Nx stands for the normal cone to Q0 at x. Since all the participants
exhibit the same behavior and want to reach the exit by the shortest path,
their spontaneous velocities can be represented as follows:
U (x) = (U0 (x1 ), . . . , U0 (xn )) with U0 (x) = −s∇D(x),
where D(x) denotes the distance between the position x = (x1 , . . . , xn ) ∈
Q0 and the exit, and where s ≥ 0 stands for the speed. By taking into account
the aforementioned postulate that in the absence of other participants each
participant tends to remain his/her spontaneous velocity until reaching the
exit, the (uncontrolled) dynamic perturbations in this model are described by
g(x) = −(s1 , . . . , sn ) ∈ Rn f or all x = (x1 , . . . , xn ) ∈ Q0 ,
where si denotes the speed of the participant i ∈ {1, . . . , n}. To control
the actual speed of all the participants in the presence of the nonoverlap-
ping condition, we suggest in [11] to involve control functions u(·) =
(u1 (·), . . . , un (·)) into perturbations as follows:
g(x(t), u(t)) = (s1 u1 (t), . . . , sn un (t)), t ∈ [0, T ].
To optimize the sweeping dynamics by using controls u(·), consider the cost
functional
 T
1 2
minimize J[x, u] := (x(T ) + u(t)2 dt) (2.34)
2 0

the meaning of which is to minimize the distance of all the participants to

the exit at the origin together with minimizing the energy of feasible controls
u(·).
The described sweeping optimal control problem falls into the category
of dynamic optimization problems with controlled dynamics, which were
considered in Section 4. The necessary optimality conditions for such prob-
lems obtained in Theorem 2.31 led us in [11] to developing an efficient
algorithmic procedure to determine optimal controls and trajectories in the
general case of finitely many participants and solve the problem analytically
in the cases where n = 2, 3. More recent results in this direction allow us to
do calculations for a fairly large number of participants in practical settings
that appear in real applications.
In the planar version of the microscopic crowd motion model [49], the
overlapping condition is not polyhedral anymore while being represented by
Q := {x ∈ R2n | Dij (x) ≥ 0 f or all i/= j},
 2.6 Conclusion 55

where Dij (x) := xi − xj  − 2R is the signed distance between the disks i
and j of the same radius R identified with n ≥ 2 participants on the plane.
The corresponding optimal control problem formulated and investigated in
[12] is described via the sweeping dynamics as follows: minimize the cost
functional (2.34) over the controlled sweeping process

−ẋ(t) ∈ N (x(t); C(t)) + g(x(t), u(t)) f or a.e. t ∈ [0, T ],
C(t) := C + v̄(t), v̄(t) = r ∈ [r1 , r2 ] on [0, T ], x(0) = x0 ∈ C(0),

where the initial data and constraints are given by

g(x(t), u(t)) := (s1 u1 (t) cos θ1 (t), s1 u1 (t) sin θ1 (t), . . . , sn un (t) cos θn (t),
 
r r
sn un (t) sin θn (t)), v̄i+1 (t) = v̄i (t) := √ , √ , i = 1, . . . , n − 1,
2n 2n
C := {x = (x1 , . . . , xn ) ∈ R2n | hij (x) ≥ 0 f or all i/= j as
i, j = 1, . . . , n}

with the functions hij (x) := Dij (x) = xi − xj  − 2R and with

x(t) − v̄(t) ∈ C f or all t ∈ [0, T ].

Observe that the sweeping dynamic in the described planar version of the
crowd motion model is generated by the set C, which is not a convex
polyhedron, while belonging the class of prox-regular sets in the sense of
[45]. The required extensions of necessary optimality conditions for this
type of controlled sweeping processes were obtained in [12] by using the
corresponding modification of the method of discrete approximations. The
obtained results were applied therein to establishing efficient relationships
which allowed us to determine optimal parameters, to develop numerical
algorithms, and realize them in particular settings.

2.6 Conclusion
This chapter discusses recent applications of advanced constructions of
variational analysis and generalized differentiation to challenging classes of
optimal control problems governed by discontinuous differential inclusions
that are known as sweeping processes. We present optimality conditions for
such control systems obtained by employing the method of discrete approxi-
mations and then review the usage of the obtained conditions to solving some
56 Applications of Variational Analysis to Controlled Sweeping Processes

practical models. There is the intensive ongoing research on further aspects

of the theory of controlled sweeping processes motivated by applications
to new models arising in nanotechnology, marine surface vehicle modeling,
electronics, etc.

Acknowledgments
This research was partially supported by the US National Science Foundation
under grants DMS-1007132 and DMS-1512846, by the US Air Force Office
of Scientific Research grant #15RT0462, and by the Australian Research
Council under Discovery Project DP-190100555.

References
[1] B. Acary, O. Bonnefon, B. Brogliato, Nonsmooth Modeling and Simu-
lation for Switched Circuits, Springer, Berlin, 2011.
[2] L. Adam, J.V. Outrata, ‘On optimal control of a sweeping process
coupled with an ordinary differential equation’, Discrete Contin. Dyn.
Syst. Ser. B 19 (2014), 2709–2738.
[3] S. Adly, T. Haddad, L. Thibault, ‘Convex sweeping process in the
framework of measure differential inclusions and evolution variational
inequalities’, Math. Program. 148 (2014), 5–47.
[4] C.E. Arroud and G. Colombo, ‘A maximum principle of the controlled
sweeping process’, Set-Valued Var. Anal. 26 (2018), 607–629.
[5] R.J. Aumann, ‘Integrals of set-valued functions’, J. Math. Anal. Appl.
12 (1965), 1–12.
[6] N.N. Bogolyubov, ‘Sur quelques méthodes nouvelles dans le Calculus
des Variations’, Ann. Math. Pura Appl. 7 (1929), 249–271.
[7] B. Brogliato, Nonsmooth Mechanics: Models, Dynamics and Control,
3rd edition, Springer, Cham, Switzerland, 2016.
[8] B. Brogliato, A. Tanwani, ‘Dynamical systems coupled with mono-
tone set-valued operators: formalisms, applications, well-posedness, and
stability’, SIAM Rev. 62 (2020), 3–129.
[9] M. Brokate, P. Krejčcí, ‘Optimal control of ODE systems involving a
rate independent variational inequality’, Discrete Contin. Dyn. Syst. Ser.
B 18 (2013), 331–348.
[10] T.H. Cao, G. Colombo, B.S. Mordukhovich, D. Nguyen, ‘Optimiza-
tion of fully controlled sweeping processes’, J. Diff. Eqs. 295 (2021),
138–186.
 References 57

[11] T.H. Cao and B S. Mordukhovich, ‘Optimal control of a perturbed

sweeping process via discrete approximations’, Discrete Contin. Dyn.
Sys. Ser. B, 21 (2016), 3331–3358.
[12] T.H. Cao, B.S. Mordukhovich, ‘Optimal control of a nonconvex per-
turbed sweeping process’, J. Diff. Eqs. 266 (2019), 1003–1050.
[13] A.A. Chikrii, Conflict Controlled Processes, Kluwer Acad. Publ.,
Boston, 1997.
[14] F.H. Clarke, Optimization and Nonsmooth Analysis, Wiley, New York,
1983.
[15] G. Colombo and P. Gidoni, ‘On the optimal control of rate-independent
soft crawlers’, J. Math. Pure Appl. 146 (2021), 405–417.
[16] G. Colombo, R. Henrion, N.D. Hoang, B.S. Mordukhovich, ‘Optimal
control of the sweeping process’, Dyn. Contin. Discrete Impuls. Syst.
Ser. B 19 (2012), 117–159.
[17] G. Colombo, R. Henrion, N.D. Hoang, B.S. Mordukhovich, ‘Optimal
control of the sweeping process over polyhedral controlled sets’, J. Diff.
Eqs. 260 (2016), 3397–3447.
[18] G. Colombo, B.S. Mordukhovich, D. Nguyen, ‘Optimal control of
sweeping processes in robotics and traffic flow models’, J. Optim.
Theory Appl. 182 (2019), 439–472.
[19] G. Colombo, B.S. Mordukhovich, D. Nguyen, ‘Optimization of a per-
turbed sweeping process by discontinuous controls’, SIAM J. Control
Optim. 58 (2020), 2678–2709.
[20] M.d.R. de Pinho, M.M.A. Ferreira, G.V. Smirnov, ‘Optimal con-
trol involving sweeping processes’, Set-Valued Var. Anal. 27 (2019),
523–548.
[21] A.Y. Dubovitskii, A.A. Milyutin, ‘Extremum problems in the presence
of restrictions,’ USSR Comput. Math. Math. Phys. 5 (1965), 1–80.
[22] J.F. Edmond, L. Thibault, ‘Relaxation of an optimal control problem
involving a perturbed sweeping process’, Math. Program. 104 (2005),
347–373.
[23] R. Gabasov, F.M. Kirillova, Qualitative Theory of Optimal Processes,
Marcel Dekker, New York, 1976.
[24] R. Hedjar, M. Bounkhel, ‘Real-time obstacle avoidance for a swarm of
autonomous mobile robots’, Int. J. Adv. Robot. Syst. 11 (2014), 1–12.
[25] R. Henrion, J.V. Outrata, T. Surowiec, ‘On the coderivative of nor-
mal cone mappings to inequality systems’, Nonlinear Anal. 71 (2009),
1213–1226.
58 Applications of Variational Analysis to Controlled Sweeping Processes

[26] N.D. Hoang, B.S. Mordukhovich, ‘Extended Euler–Lagrange and

Hamiltonian formalism in optimal control of sweeping processes with
controlled sweeping sets’, J. Optim. Theory Appl. 180 (2019), 256–289.
[27] A.D. Ioffe, V.M. Tikhomirov, Theory of Extremal Problems, North-
Holland, Amsterdam, The Netherlands, 1979.
[28] Y.P. Kondratenko, A.A. Chikrii, V.F. Gubarev, J. Kacprzyk (eds.),
Advanced Control Techniques in Complex Engineering Systems: Theory
and Applications, Studies in Systems, Decision and Control, Vol. 203,
Springer, Cham, Switzerland, 2019.
[29] Y.P. Kondratenko, V.M. Kunsevich, A.A. Chikrii, V.F. Gubarev (eds.),
Advanced Control Systems: Theory and Applications, Series in Automa-
tion, Control and Robotics, River Publishers, Gistrub, Denmark, 2021.
[30] M.A. Krasnosel’skii, A.V. Pokrovskii, Systems with Hysteresis,
Springer, Berlin, 1989.
[31] N.N. Krasovskii, A.I. Subbotin, Game-Theoretical Control Problems,
Springer, New York, 1988.
[32] V.M. Kunsevich, V.F. Gubarev, Y.P. Kondratenko, D. Lebedev, V.
Lysenko (eds.), Control Systems: Theory and Applications, Series in
Automation, Control and Robotics, River Publishers, Gistrub, Denmark,
2018.
[33] M. Kunze, M. D. P. Monteiro Marques, “An introduction to Moreau’s
sweeping process,” in: Impacts in Mechanical Systems (B. Brogliato,
ed.), Lecture Notes in Phys. 551, pp. 1–60, Springer, Berlin, 2000.
[34] A.A. Lyapunov, ‘Sur les founctions-vecteurs complétement additives’,
Izvest. Akad. Nauk SSSR, Ser. Mat. 3 (1940), 465–478.
[35] B.S. Mordukhovich, Methods of Approximations in Problems of Opti-
mization and Control, Nauka, Moscow, 1988.
[36] B.S. Mordukhovich, Variational Analysis and Generalized Differentia-
tion, I: Basic Theory, II: Applications, Springer, Berlin, 2006.
[37] B.S. Mordukhovich, Variational Analysis and Applications, Springer,
Cham, Switzerland, 2018.
[38] B.S. Mordukhovich, ‘Optimal control of Lipschitzian and discontinu-
ous differential inclusions with various applications’, Proc. Inst. Math.
Mech. Azer. Acad. Sci. 45 (2019), 52–74.
[39] B.S. Mordukhovich, R.T. Rockafellar, ‘Second-order subdifferential
calculus with applications to tilt stability in optimization’, SIAM J.
Optim. 22 (2012), 953–986.
[40] J.J. Moreau, ‘Ra e par un convexe variable, I’, Trav. Semin. d’Anal.
Convexe, Montpellier 1 (1971), Exposé 15, 36 pp.
 References 59

[41] J.J. Moreau, ‘On unilateral constraints, friction and plasticity’, in New
Variational Techniques in Mathematical Physics (G. Capriz and G.
Stampacchia, eds.), Proc. C.I.M.E. Summer Schools, pp. 173–322,
Cremonese, Rome, 1974.
[42] L.S. Pontryagin, V.G. Boltyanskii, R.V. Gamkrelidze, E.F. Mishchenko,
The Mathematical Theory of Optimal Processes, Wiley, New York,
1962.
[43] B.N. Pschenichnyi, Convex Analysis and Extremal Problems, Nauka,
Moscow, 1980.
[44] B.N. Pschenichnyi, V.V. Ostapenko, Differential Games, Naukova
Dumka, Kyiv, 1992.
[45] R.T. Rockafellar, R.J-B. Wets, Variational Analysis, Springer, Berlin,
1998.
[46] G. V. Smirnov, Introduction to the Theory of Differential Inclusions,
American Mathematical Society, Providence, RI, 2002.
[47] A.A. Tolstonogov, Differential Inclusions in a Banach Spaces, Kluwer,
Dordrecht, The Netherlands, 2000.
[48] A.A. Tolstonogov, ‘Control sweeping processes’, J. Convex Anal. 23
(2016), 1 099–1123.
[49] J. Venel, ‘A numerical scheme for a class of sweeping process’,
Numerische Mathematik, 118 (2011), 451–484.
[50] R.B. Vinter, Optimal Control, Birkhaüser, Boston 2000.
[51] L.C. Young, ‘Generalized curves and the existence of an attained
absolute minimum in the calculus of variations’, C. R. Soc. Sci. Lett.
Varsovie, Cl. III, 30 (1937), 212–234.
[52] V. Zeidan, C. Nour, H. Saoud, ‘A nonsmooth maximum principle for
a controlled nonconvex sweeping process’, J. Diff. Eqs. 269 (2020),
9531–9582.
 3
Robust and Robustly-Adaptive Control
of Some Noninvertible Memoryless Systems

L. Zhiteckii1 and K. Solovchuk2

1 InternationalResearch and Training Center for Information Technologies
and Systems of the National Academy of Science of Ukraine and Ministry
of Education and Sciences of Ukraine,
Acad. Glushkova av., 40, Kyiv, 03187, Ukraine
2 Poltava Scientific Research Forensic Center of the MIA of Ukraine,

Rybalskiy lane, 8, Poltava, 36011, Ukraine

E-Mail: [email protected]; [email protected]

Abstract
This chapter deals with the robust control of some uncertain multivariable
memoryless (static) discrete-time systems in the presence of arbitrary unmea-
surable bounded disturbances, whose bounds may be unknown, in general.
The main feature of the systems to be controlled is that their gain matrices
are noninvertible. Again, they are of not full rank. It is assumed that the
elements of these matrices are exactly unknown. However, the bounded
intervals to which they belong are a priori known. The nonadaptive robust
control approach applicable to the case when the lengths of these intervals
are relatively short is proposed. Such an approach is based on the so-called
pseudoinverse model-based concept. The robustly-adaptive control utilizing
a modification of inverse-model concept is designed to cope with large
parametric uncertainty. The asymptotic properties of both nonadaptive and
adaptive feedback control systems are derived. To support the theoretic study,
numerical examples and simulation results are given.

61
62 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

Keywords: Boundedness, discrete time, estimation algorithm, feedback con-

trol system, robust adaptive controller, pseudoinverse model-based control,
uncertainty.

3.1 Introduction
The problem of a perfect performance of control systems is an important
problem from both theoretical and practical points of view. Within the frame-
work of this actual problem, new approaches have been advanced by many
researches. The latest results in this scientific area have been reported in
numerous papers and generalized in several books including [1, 2]. Among
them, the recent works [3–7] dealt with advanced multivariable control
systems.
A long-standing problem of the optimal controller design for multivari-
able system in the presence of unmeasurable disturbances [8] has been solved
using, in particular, the internal model control principle proposed in [9].
A perspective modification of this principle is the so-called model-inverse
approach. Since the pioneering work [10], the problem of inversion of linear
time-invariant multivariable systems has attracted an attention of several
authors [11–14]. Recently, a significant progress in this research are has been
achieved in [15, 16].
An inverse model-based approach to ensuring a perfect performance of
linear multivariable control systems containing the memoryless (static) plants
was first advanced in [17]. Regardless of this work, a similar approach
was proposed in [18] to deal with some memoryless plants. However,
the approach above mentioned is quite unacceptable if their gain matrices
are either square but singular or nonsquare because they are noninvertible
matrices.
It turned out that so-called pseudoinverse (generalized inverse) model-
based concept advanced in the paper [19] can be exploited to cope with the
possible noninvertibility of gain matrices. This fruitful concept was extended
in [20, 21] to robust control of a noninvertible and uncertain plant with
unmeasurable bounded disturbances.
The robust control theory gives a powerful tool to reject the unmea-
sured disturbances in the closed-loop control system with parametric and
nonparametric uncertainty. The books [22–27] et al. provide insight into the
results achieved in this theory to the beginning of the 2000s. A new esti-
mation method applicable to a nonlinear discrete-time system with bounded
disturbances has been advanced in the work [28]. Novel results with respect
 3.2 Problem Statement 63

to a practical application of the robustness theory is reported, in particular, in

recent paper [29].
Unfortunately, the robust control theory may not be employed if the initial
parametric uncertainty is “wide” enough. Meanwhile, the adaptive approach
gives some universal tool to deal with such a type of uncertainty. Foundations
of this approach have been extended and generalized in the books [26, 30–35].
Difficulties that take place when adaptive control uses the point estimation
algorithms are how to guarantee the stability (the boundedness) of the closed-
loop control system [30, 31]. To overcome these difficulties, the so-called
Frequency theorem given in [30, theorem 4.Π.3] or the Key Technical Lemma
given in [31, subsect. 6.2] are usually utilized to establish the boundedness
properties of adaptive control systems. However, these tools become not
applicable if the gain matrix is noninvertible; see [30, p. 242] and [31, p. 202].
Nevertheless, the problem of the adaptive control of nonsquare multivariable
memoryless plants whose gain matrices have the full rank has been solved in
the work [36].
This chapter extends the results of [21] concerning the robust nonadap-
tive approach to controlling uncertain noninvertible memoryless systems
whose gain matrices have not full rank and generalizes the robustly-adaptive
techniques proposed in [37 – 40] to deal with these systems.

3.2 Problem Statement

Let
yn = Bun−1 + vn (3.1)
be the vector-valued difference equation of a static (memoryless) plant that
is some linear multivariable discrete-time system to be stabilized. In this
equation, yn ∈ Rm , un ∈ Rr and vn ∈ Rm represent the measured output,
control input and unmeasured external disturbance vectors, respectively, at
(1) (m)
the nth time instant (n = 1, 2, . . . ) defined by yn = [yn , . . . , yn ]T ,
(1) (r) (1) (m)
un = [un , . . . , un ]T and vn = [vn , . . . , vn ]T ,
⎛ (11) ⎞
b . . . b(1r)
B = ⎝ ··· ··· ··· ⎠ (3.2)
b(m1) . . . b(mr)
is an arbitrary time-invariant m × r gain matrix.
(1) (r)
The case when the number of the control inputs un , . . . , un is not
(1) (m)
less than two but does not exceed the number of the outputs yn , . . . , yn
64 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

meaning that
2 ≤r≤m (3.3)
is here considered. However, in contrast with problem stated in the work [36],
the rank of B satisfies the strict inequality
rank B<r (3.4)
implying that B is a not full rank matrix.
The following basic assumptions with respect to the gain matrix B and
(i) (i)
the sequences {v (i)
n } = v0 , v1 , . . . (i = 1, . . . , m) are introduced.
A1) The elements of B are all unknown. However, there are some interval
estimates defined as
(ij)
b(ij) ≤b(ij) ≤b , i= 1, . . . , m, j= 1, . . . , r (3.5)
(ij)
where the upper and lower bounds b(ij) and b , respectively, are
assumed to be known.
A2) The rank of B satisfies Equation (3.4) giving that B is a noninvertible
matrix.
(i)
A3) vn s (i = 1, . . . , m) are all the arbitrary scalar variables bounded in
modulus according to & &
& (i) &
&vn & ≤ε(i) <∞

where ε(i) s are constant. It is assumed that these upper bounds may be
unknown, in general.
Let y 0 = [y 0(1) , . . . , y 0(m) ]T be a &desired & output vector
& 0(m) & (y 0(i) ≡
const ∀i= 1, . . . , m). Suppose that &y & +· · ·+ &y
0(1) & = 0 implying
that, at least, one y 0(i) 0(1)
of y , . . . , y 0(m) is nonzero.
Define the output error vector
en = y 0 − yn (3.6)
(i) (i) (1) (m)
with the components en = y 0(i) − yn , i.e., en = [en , . . . , en ]T .
The problem is to design the feedback controller guaranteeing the
ultimate boundedness of the sequence {en } = e1 , e2 , . . . , meaning
lim sup ||en || <∞ (3.7)
n→∞

provided
lim sup ||un || <∞ (3.8)
n→∞
 3.3 Preliminaries 65

Remark 3.1. Assumptions A1) – A3) differ from similar assumptions made
in [36] only in that B is here the matrix of not full rank, and it is its essential
feature.
Remark 3.2. Equation (3.8) is here introduced additionally since it may not
be satisfied even if (3.7) takes place.

3.3 Preliminaries
Assume, for the time being, that B is a known noninvertible matrix of full
rank (rank B = r), i.e., B represents some nonsquare matrix with r < m. In
this case, the so-called pseudoinverse control

un = un−1 + B + en (3.9)

proposed by the authors in their previous works (see, e.g., [21]) ensures
the minimum of the upper bound on the Euclidean norm ||en ||2 of the
output error vector of the closed-loop control system (3.1), (3.6), (3.9)
with any bounded sequence {vn } = v1 , v2 , . . . Here the notation P +
of any pseudoinverse (generalized inverse) matrix introduced in [41] and
defined as
P + = lim (P T P + δIr )−1 P T (3.10)
δ→0
where Ir denotes the identity r × r matrix is used. Recall that if rank P = r
for some P ∈ Rm×r , then the expression of P + is simplified [42, item 7.46]
instead of (3.10) we have

P + = (P T P )−1 P T

This control system is designed as shown in Figure 3.1.

Figure 3.1 Configuration of the pseudoinverse model-based feedback control system

without uncertainty
66 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

Figure 3.2 Configuration of the pseudoinverse model-based feedback control system with
parametric uncertainty

In this control system, the variable ∇un := un − un−1 produced by the

pseudoinverse model represents the increment of the control action during
one step determined as
∇un := B + en

whereas the signal un is formed as the sum un = nk=1 ∇uk .
Since B is indeed unknown, it needs to be replaced by an appropriate
(ij)
estimate B0 = (b0 ), i = 1, . . . , m, j = 1, . . . , r, that is a fixed
matrix (in the nonadaptive case) with rank B0 = r taken from the interval
matrix set
' (
Ξ = {(b(ij) ) : b(ij) ∈ b(ij) , b̄(ij) , i = 1, . . . , m, j = 1, . . . , r}
(3.11)
of admissible matrices B satisfying the constraints (3.5). Then, under the
parameter uncertainty given as B ∈ Ξ, instead of (3.9), the control law
becomes
un = un−1 + B0+ en (3.12)
Figure 3.2 represents the structure of the feedback control system contain-
ing the pseudoinverse model-based controller described by Equations (3.6),
(3.12).
Introduce the matrix
Δ := B0 − B
(ij) (ij) (ij)
which, by virtue of the fact that b(ij) ≤ b(ij) ≤ b and b(ij) ≤ b0 ≤ b ,
is specified as
  ' (
Δ = δ (ij) : δ (ij) ∈ δ (ij) , δ̄ (ij)

(ij) (ij)
where δ (ij) = b0 − b̄(ij) and δ̄ (ij) = b0 − b(ij) .
 3.3 Preliminaries 67
(i)
Recall that if there are no disturbances (vn ≡ 0 for all i = 1, . . . , m),
then an equilibrium state of the closed-loop control system (3.1), (3.6), (3.12)
defined by the pair (ue , y e ) in which y e = Bue , where ue represents the
solution of the equation
B0+ Bu = B0+ y 0 (3.13)
with respect to u, exist provided that rank B0 = r and the condition

q<1 (3.14)

where && + &&

q := max &&B Δ&& (3.15)
(ij)
0
Δ : δ (ij) ∈ [δ , δ̄ (ij) ]

is satisfied [36, Lemma 7.1]. In this lemma it has been established that the
condition (3.14) together with (3.15) is sufficient to achieve the ultimate
boundedness of the sequences {un } and {en } if B is a matrix of full rank.
To establish the sufficient conditions under which the equilibrium state
exist, the requirement that the r × r matrix B0+ B of Equation (3.13) is
nonsingular is utilized. ) *
It turns out that when y 0 ∈
/ ℵ B0+ and also y 0 ∈/ R (B) , where ℵ (P )
and R (P ) denote the null space and the range of a matrix P, respectively,
then the equilibrium state may exist even if det B0+ B = 0. To show this,
consider an example.
Example 3.1. Let
⎛ ⎞ ⎛ ⎞
1 −2 3 −4.5
B = ⎝ 2 −4 ⎠ , B 0 = ⎝ 1 −1.5 ⎠
−1 2 −2 3

be some nonsquare matrices of the not full rank (rank B = rank B0 = 1).
By (3.10) we have
 
+ 1 6 2 −4
B0 =
91 −9 −3 6
) *
Put y 0 = [1, 3, 5]T . Such a choice of y 0 gives y 0 ∈
/ ℵ B0+ and y 0 ∈ /
R (B). It can be verified that in this case, Equations (3.13) will have the
infinite number of solutions ue = [ue(1) , ue(2) ]T lying on the straight line
described by
7ue(1) − 14ue(2) + 4 = 0
68 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

Figure 3.3 Set Eu of equilibrium points in Example 3.1

(see Figure 3.3). In other words, the set Eu = {ue } of the equilibrium points
becomes here infinite.
Unfortunately, if rank B < r and y 0 ∈ / R (B), then the equilibrium state
may not exist, in principle. To demonstrate this annoying fact, consider the
following numerical example.
Example 3.2. Suppose B and B0+ are some square matrices given by
   
2 4 + 1 −2
B= , B0 =
1 2 2 −4
) *
and y 0 ∈/ ℵ B0+ , y 0 ∈ / R (B). Since B0+ B = 0r×r , in this case, Equation
(3.13) has no solutions. This example shows that the closed-loop control
system defined by Equations (3.1), (3.6) and (3.12) will have no equilibrium
state.
It turns out that the equilibrium state {ue , y e } of the control system above
mentioned exists iff
) * ) *
ℵ B0+ ∩ y 0 + R (B) = ∅ (3.16)

where the symbol “+” denoted the Minkowski sum of two sets of vectors
in Euclidean space. Actually, by the definition of set R (B) to which the
zero vector 0m belongs, follows that (−Bu) ∈ R (B). Thereby, the vector
y 0 − Bu belongs to the set y 0 + R (B) representing a linear manifold whose
dimension is defined as dim R (B) = rank B [41, p. 38]. Geometrically, this
set is parallel to the linear subspace R (B) and does not contain the vector 0m .
 3.4 Robust Nonadaptive Control 69

Figure 3.4 Geometric interpretation of the condition

) *
On the other hand, the vector y 0 − Bu should belong to the set ℵ B0+ to
satisfy Equation (3.13). Therefore, the )equilibrium
* point ue exist if and only
if the intersection of y + R (B) and ℵ B0 is a non-empty set according to
0 +

the condition defined by Equation (3.16).

Figure 3.4 gives the geometric interpretation of the condition (3.16).
Unfortunately, the condition (3.16) is not constructive. Another condi-
tions which seems to be constructive are established in the next section.

3.4 Robust Nonadaptive Control

Before going to the design of the robust nonadaptive control of systems whose
gain matrices are not full rank, we first establish conditions guaranteeing the
existence of the equilibrium state of the closed-loop control systems shown
in Figure 2 with rank B0 < r. To do this, as the first we show that

rank P + = rank P (3.17)

regardless of the rank of P. Indeed, according to [42, item 6.25] we)have*

rank P = dim R (P ). On the other hand, it is known that R (P ) =)R *P T
[41, item 3]. This gives rank P + = dim R (P + ) = dim R P T =
rank P T = rank P leading to (3.17).
70 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

Now, fix a m×r matrix B0 whose rank is equal 1. Due to Equation (3.17),

we have rank B0+ = rank B0 = 1. Such a choice of B0 gives that this matrix
takes the following form:
⎛ (11) (11) (11)
⎞
β0 k12 β0 . . . km−1, m β0
⎜ ⎟
⎜ β (21) k β (21) . . . k (21) ⎟
⎜ 12 m−1, m β ⎟
B0+ = ⎜ ⎟
0 0 0
⎜ .. .. .. .. ⎟ (3.18)
⎜ . . . . ⎟
⎝ ⎠
(r1) (r1) (r1)
β0 k12 β0 . . . km−1, m β0
- ./ 0
m columns

(Its columns are collinear r-dimensional vectors.)

Multiplying B0 given in the form (3.18) by B of the form (3.2) we obtain
the matrix ⎛ (11) ⎞
g . . . g (1r)
⎜ .. .. ⎟
G = B0+ B = ⎝ ... . . ⎠ (3.19)
g (r1) . . . g (rr)
that is the matrix of Equation (3.13). In this expression, g (1k) = g (1k) (B) =
(11) (r1)
β0 dk , . . . , g (rk) = g (rk) (B) = β0 dk denote the elements of the kth
column of G where

dk = b(1k) + k12 b(2k) + . . . + km−1,m b(mk) (3.20)

is the scalar variable which depends on the elements of its kth column.
Introduce the r-dimensional vector
(11) (r1) T
g (k) (B) = [β0 , . . . , β0 ] dk (3.21)

which is the kth column of G = G (B) depending on B for a fixed B0 . Since

Ξ defined by Equation (3.11) is a bounded closed set, the minimum

dk (B) = min dk (B)

B∈Ξ

and the maximum

dk (B) = max dk (B)
B∈Ξ

of the variable dk (B) exist. Taking into account this fact and also Equation
(3.20) it can be written
 3.4 Robust Nonadaptive Control 71
 
dk (B) = min b(1k) + k12 b(2k) + · · · + km−1,m b(mk)
(ik)
b(ik) ∈ [b(ik) , b ]
i = 1, . . . , m
k = 1, . . . , r (3.22)

dk (B) = max (b(1k) + k12 b(2k) + · · · + km−1, m b(mk) )

(ik)
b(ik) ∈ [b(ik) , b ]
i = 1, . . . , m
k = 1, . . . , r) (3.23)

Let the condition

dk (B) dk (B) > 0 (3.24)
be satisfied for, at least, one k ∈ 1, . . . , r.
This condition implies that dk (B) and dk (B) are nonzero and the same
sign. Since they are linear continual functions of the m elements of B
(according to Equation (3.20)), all components of g (k) (B) are nonzeros for
all possible B from Ξ, if the condition (3.24) is satisfied. The transposition of
G = G (B) leads to the matrix
⎛ (11) (r1) ⎞
β0 d1 . . . β0 d1
⎜ .. .. .. ⎟
GT = ⎝ . . . ⎠ (3.25)
(11) (r1)
β0 dr . . . β 0 dr

non-zero elements of which are collinear vectors. By the definition of the null
space ℵ (P ) = {x : P x = 0r } of any r × r matrix P (see, e.g. ,[42, item
6.24]) and by virtue of the expression (3.25) it can be written
(1)T
ℵ(GT (B)) = {u : β0 u = 0r } (3.26)

where
(1) (11) (r1) T
β0 = [β0 , . . . , β0 ] (3.27)
Since the dimension of ℵ(GT (B)) is equal to dim ℵ(GT (B)) = r −
rank GT (B) = r − 1 (according to [42, item 6.25]), it is not hard to see
that, for each B, the set ℵ(GT (B)) is a hyperplane in the space Rr of some
(1)
vectors u to which β0 given by the expression (3.27) is the normal vector.
72 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

Next, consider the right-hand side of Equation (3.13). Substituting the

expression (3.18) into this side together with y 0 = [y 0(1) , . . . , y 0(m) ]T and
taking the expression (3.27) into account, we get
(1) T
B0+ y 0 = β0 L (3.28)

where L = y 0(1) + k12 y 0(2) + · · · + km−1,m y 0(m) .

Equation (3.28) shows that when L = 0, then the right-hand side of
Equation (3.13) becomes a nonzero r-dimensional vector orthogonal to the
null space of GT (B) (in accordance with (3.26)):
B0+ y 0 ⊥ℵ(GT (B)) (3.29)
If L = 0 and the condition (3.29) is satisfied, then Equation (3.13) will be
compatible (see [42, item 6.34]). On the other hand, if L = 0, then Equation
(3.13) becomes homogeneous; it has always a solution.
The fact established above leads to the following basic result.
Lemma 3.1. Fix a matrix B0 with rank B0 = 1 so that all the elements of
B0+ are nonzero. If the condition (3.24) with dk (B) and dk (B) defined by
Equations (3.22) and (3.23), respectively, are satisfied, at least, for one k ∈
1, . . . , r, then the closed-loop control system (3.1), (3.6), (3.12) without any
disturbances has an equilibrium state for any B from Ξ and for arbitrary y 0 ∈
Rm .
The geometric interpretation of the property (3.29) written in the form
B0+ y 0 ⊥ℵ(B0+ B )
- ./ 0
B∈Ξ

is given in Figure 3.5.

Remark 3.3. It is not hard to see that the computation of dk (B) and
dk (B) by Equations (3.22) and (3.23), respectively, leads to solving
the standard linear programming problem (LP): find min dk (B) and
(ij)
max dk (B) (k = 1, . . . , r) in the presence of constraints b(ij) ≤b(ij) ≤b
( i= 1, . . . , m, j= 1, . . . , r).
With these preliminary results in mind, the asymptotic properties of
the closed-loop control system containing the pseudoinverse model-based
controller (3.12) are finally established in theorem below.
Theorem 3.1. Subject to Assumptions A1) to A3), if the conditions of Lemma
3.1 added by the condition (3.14) together with the expression (3.15) are
 3.4 Robust Nonadaptive Control 73

Figure 3.5 Geometric interpretation of the condition (3.29)

satisfied, then the control law (3.12), (3.6) when applied to the plant (3.1)
is robust with respect to the uncertainty (3.5), and the ultimate boundedness
of the sequences {en } and {un } is guaranteed:
 
lim sup ||en || = ||Im − B0 B0 || ||e0 || + 2 sup ||vn ||
+
n→∞ 0<=n<∞

+2 sup ||vn ||(1 − q)−1 < ∞

0=n<∞

lim sup ||un − ue || = ||Ir − B0+ B0 || ||u0 − ue ||

n→∞
 
+||B0+ || (1 − q)−1 sup ||vn || < ∞
0<=n<∞

Proof. See [21]. 

Remark 3.4. Note that B0 may not belong to Ξ to satisfy rank B0 = 1 as
illustrated in Figure 3.6.
At first sight, it is
' hard to verify
( the condition (3.14) by looking over all
(ij) (ij) (ij)
possible δ s from δ , δ to compute the value of q (according to the
expression (3.15)). In fact, it is not the case. The point is that condition (3.14)
74 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

Figure 3.6 Choice of matrix B0 satisfying the condition rank B0 = 1 (possible cases): (a)
there are B0 s belonging to Ξ; (b) there are no B0 s belonging to Ξ (3.29)

holds regardless of how the norm ||  || is chosen in the expression (3.15).

Therefore, the computation difficulty above mentioned can be avoided if we
will choose this norm as ||B0+ Δ||1 , where

!
s
||P ||1 := max |p(ij) | (3.30)
1≤i≤s
j=1

denotes the so-called 1-norm of a s×s matrix P. (According to the expression

(3.30) this norm represents a piecewise-linear function of δ (ij) ). With this in
mind, we can consider the maximization problems caused by the expression
(3.15) as a LP problems stated as follows.
Find
min σ (ij) max σ (ij) (3.31)
under constraints
(ij)
δ (ij) ≤ δ (ij) ≤ δ i = 1, . . . , m, j = 1, . . . , r (3.32)

where
!
m
(ik) (kj)
σ (ij) = β0 δ
k=1
are the linearforms, depending on the elements of the pseudoinverse matrix
(ij) (ij)
B0+ = β0 , and δ (ij) and δ are defined by the expression (3.32).
 3.4 Robust Nonadaptive Control 75

Such a method allows us to reformulate Theorem 3.1 in the terms of LP

problems as follows.
Theorem 3.2. Let the conditions of Lemma 3.1 be satisfied. Then if
!
r
max{ |min σ (ij) |, |max σ (ij) |} < 1 ∀i = 1, . . . , m
j=1

with min σ (ij) and max σ (ij) that are the solutions of LP problems given
in the expression (3.31), then the controller described in Equations (3.12)
together with (3.6) robustly stabilizes the system (3.1) for any B from the set
Ξ defined by the expression (3.11).
Proof. Follows immediately from the definition (3.30) of the 1-norm of an
arbitrary matrix. 
Example 3.3. Let Ξ = {(b(ij) ) : 0.4 ≤ b(11) ≤ 1.4, −2 ≤ b(12) ≤
−1, −1.2 ≤ b(21) ≤ −0.2, 0.7 ≤ b(22) ≤ 1.7, −1.7 ≤ b(31) ≤ −0.7, 1 <=
b(32) ≤ 2}. Next, choose
⎛ ⎞
0.8 −1.6
B0 = ⎝ −0.7 1.4 ⎠
−1.0 2.0
with rank B0 = 1. Such a choice yields
 
+ 16/213 −14/213 −20/213
B0 =
−32/213 28/213 40/213

Further, according to (3.18), we get k12 = −7/8, k23 = −5/4. By the

formulas (3.22) and (3.23), it can be found d1 (B) = 1.45, d1 (B) = 4.575
showing that d1 (B) d1 (B) > 0.
Since the conditions of Lemma 3.1 are here satisfied, we conclude that
the closed-loop control system described by Equations (3.1), (3.12) together
with (3.6) has the equilibrium state for any B ∈ Ξ and arbitrary y 0 ∈ Rm .
Let b(11) = 1, b(12) = −1.5, b(21) = −0.8, b(22) = 1.2, b(31) = −1.2,
b(32) = 1.8. Choose y 0 = [0.3, 0.4, 0.3]T to satisfy y 0 ∈
/ R (B). Substituting
0
the expressions of B, B0 and y into Equation (3.13) gives
3.2ue(1) − 4.8ue(2) + 0.425 = 0
This equation describes the set Eu = {ue } of the vector ue = [ue(1) , ue(2) ]T
which is not one-point set. This set is shown in Figure 3.7.
76 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

Figure 3.7 Equilibrium states related to the conditions of Example 3.3

Solving the LP problems (3.31) yields minσ (11) = −194/2130,

maxσ (11) = 306/2130, minσ (12) = −106/2130, maxσ (12) =
394/2130, minσ (21) = −612/2130, maxσ (21) = 388/2130, minσ (22) =
−788/2130, maxs(22) = 216/2130. This gives

!
2
max{ |min σ (1j) |, |max σ (1j) |} = 70/213 < 1,
j=1

!
2
max{ |min σ (2j) |, |max σ (2j) |} = 140/213 < 1.
j=1

Thus, all the conditions of Theorem 3.2 are here satisfied.

Simulation experiment 3.1. To verify how the pseudoinverse controller with
fixed B0+ , before chosen, works, a simulation of the control system with given
(1)
B, B0 and y 0 was conducted. The sequences {vn } and {v (2) n } are generated
as some i.i.d. random variables belonging to [−0.07, 0.07]. Duration of this
simulation experiment was equal to 100 steps.
Simulation results are given in Figure 3.8. They demonstrate the bound-
edness of ||un ||2 and ||yn ||2 .

3.5 Robustly-Adaptive Control of Square Systems (Case 1)

In this section, the robustly-adaptive control system in the case when the
(i)
bounds on the disturbances vn are known a priopi is designed. Let B be
 3.5 Robustly-Adaptive Control of Square Systems (Case 1) 77

Figure 3.8 Robust pseudoinverse model-based controller: (a) Euclidean norm ||un ||2 of
control input vectors; (b) Euclidean norms ||yn ||2 , ||y 0 ||2 of output vectors (solid line) and
desired output vectors (dashed line), respectively

an unknown square singular r × r matrix, i.e.,

det B = 0 (3.33)
Basic idea to deal with a matrix B satisfying the requirement (3.33)
is to replace adaptive identification of the true plant having the singular
gain matrix B to the adaptive identification of a fictitious plant with the
nonsingular gain matrix B̃ of the form
B̃ = B + δ0 Ir (3.34)
where Ir denotes the identity r × r matrix and δ0 is a fixed quantity [37].
Although B̃ as well as B remain unknown, the requirement
det B̃ = 0 (3.35)
can always be satisfied by the suitable choice of δ0 in Equation (3.34). In
fact, each ith eigenvalue λi (B) of B lies in one of the r closed regions of the
complex z-plane consisting of all the Gerŝgorin discs [43, p. 146]:
& & !r & &
& & & (ij) &
&z − b(ii) & ≤ &b &, i = 1, . . . , r (3.36)
j=1
j = i
78 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

Since, at least, one of the eigenvalues λi (B) is equal to zero (due to the
singularity of B), by virtue of Equation (3.35) there are the numbers
!
r & & !
r & &
& (ij) & (i) & (ij) &
β (i) := b(ii) − &b &, β := b(ii) + &b & (3.37)
j=1 j=1
j = i j = i

such that if
|b(i1) | + · · · + |b(ir) | = 0 (3.38)
(i) (i)
then either β (i) ≤ 0 but β > 0 or β (i) < 0 but β ≥ 0. These numbers are
defined by the intersection of the ith Gerŝgorin disc with the real axis of the
complex z-plane as shown in Figures 3.9 and 3.10, respectively, left. In both
(i) (i)
cases, β (i) β ≤ 0 if the inequality (3.38) is satisfied because β (i) and β
cannot have the same sign.

(1)
Figure 3.9 The Gerŝgorin discs for r = 2 in the case |β (2) | < |β |

(2)
Figure 3.10 The Gerŝgorin discs for r=2 in the case |β | < |β (1) |
 3.5 Robustly-Adaptive Control of Square Systems (Case 1) 79

Denoting
(1) (r)
β := min{ β (1) , . . . , β (r) }, β := max{ β ,...,β }, (3.39)

consider the following two cases: (a) |β| < |β|; (b) |β| > |β| (The case
when |β| = |β| can be combined with any two cases.) In order to go to the
gain matrix B̃ of the fictitious plant having the form (3.34) in the case (a), it
is sufficient to shift the Gerŝgorin disc (3.36) right taking

δ0 > |β| (3.40)

as shown in Figure 3.9, right. In the case b), the discs (3.36) need to be shifted
left according to
δ0 < −|β| (3.41)
See Figure 3.10, right. In both cases, the nonsingularity of B̃ is guaran-
teed. Nevertheless, the conditions (3.40) and (3.41) cannot be satisfied, as yet.
In fact, the numbers β and β given by the expressions (3.39) depend of β (i)
(i)
and β s defined by Equations (3.37). But they are unknown because b(ij) s
are all unknown.
The following operations are proposed to choose a number d0 satisfying
the reqiurement (3.35). Introduce
!
r
(ij)
β (i)
min
:=b(ii) − max{ |b(ij) |, |b |}
j=1
j = i
(i) (ii) !
r
(ij)
β max := b + max{ |b(ij) |, |b |} (3.42)
j=1
j = i

(ij)
minimizing and maximizing in b(ij) ∈ [b(ij) , b ] the right-hand sides of
(i)
Equations (3.37) for β (i) and β , respectively.
Further, the number δ0 is found to satisfy the conditions

δ0 > −β min if |β min | < |β max |

δ0 > β max if |β min | > |β max | (3.43)
80 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

where β min , β max represent some quantities defined as follows:

β min := {β (1)
min
, . . . , β (r)
min
}
(1) (r)
β max := {β max , ..., β max } (3.44)

It can be clarified that if the conditions (3.43) together with Equations (3.42)
and (3.44) will be satisfied then the condition (3.35) will without fail be
ensured.
After determining the quantity δ0 we can proceed to the consideration of
(1) (r)
the fictitious plant. Since the input variables un , . . . , un and the distur-
(1) (N )
bances vn , . . . , vn of both true plant and fictitious plant are the same, this
feature makes it possible to describe our fictitious plant by the equation

ỹn = B̃un−1 + vn , (3.45)

(1) (r)
similar to Equation (3.1), where ỹn = [ỹn , . . . , ỹn ]T denotes the output
vector of the fictitious plant.
It is interesting that the components of ỹn can be measured while the com-
ponents of vn in Equation (3.45) remain unmeasurable. In fact, substituting
the expression (3.34) into Equation (3.45) due to Equation (3.1) we produce

ỹn = yn + δ0 un−1 . (3.46)

It is seen from Equation (3.46) that ỹn can always be found indirectly having
un and yn to be measured.
Now, our problem reduces to the known problem of adaptive control
applicable to the fictitious plant (3.45) with the unknown gain matrix B̃ in the
(1) (r)
presence of arbitrary bounded disturbances vn , . . . , vn . Its solving follows
the steps of the section above. Namely, the adaptive control law is designed
in the form
un = un−1 + B̃n−1 ẽn , (3.47)
in which, instead of the current estimate Bn of B, another B̃n is exploited,
and the error vector en defined in Equation (3.6) is replaced by

ẽn = y 0 − ỹn (3.48)

with ỹn given by the expression (3.46).

 3.5 Robustly-Adaptive Control of Square Systems (Case 1) 81

The adaptive identification algorithm used to determine the estimates B̃n

may be taken as
⎧
∗(i)
⎪ (i)
⎨ b̃n−1 if |ẽn | = ε0i ,
b̃(i)
n =
⎪ ∗(i) ∗(i)
⎩ b̃(i) + γn(i) ẽn −ε̂i sign ẽ2n ∇un−1 otherwise, i= 1,. . ., r
n−1 ||∇un−1 ||2
(3.49)
which is similar to that in [36]. In this algorithm, ε0i and εi are given by

ε0i > εi = 2εi , i = 1, . . . , r (3.50)

ẽ∗(i)
(i)T
n = ∇ỹn(i) − b̃n−1 ∇un−1 (3.51)
represent the ith component of the identification error ẽ∗n given as

ẽ∗n = ∇ỹn − B̃n−1 ∇un−1 (3.52)

' (
(i) (i) (i) (i)T (i1) (ir)
where ∇ỹn := ỹn − ỹn−1 , and the notation b̃n := b̃n , . . . , b̃n of
(i)
the ith row of B̃n is introduced. The coefficients γn s are chosen to satisfy

det B̃n = 0

The feedback adaptive robust control system described in Equations (3.1),

(3.47), (3.49) is designed as depicted in Figure 3.11. In this figure, the
notation ∇ỹn∗ := B̃n−1 ∇un−1 is introduced.
The asymptotic properties of the adaptive control system are established
in the theorem below.
Theorem 3.3. Determine δ0 using the formula (3.43) together with Equations
(3.42) and (3.44), and choose an arbitrary initial B̃0 = B0 + δ0 I with B0 =
(ij) (ij) (ij)
(b0 ) whose elements satisfy the conditions b(ij) ≤ b0 ≤ b . Subject
to assumptions A1–A3, the adaptive controller described in Equations (3.47)
and (3.49) together with Equations (3.46) and (3.48) when applied to the plant
(3.1) yields Equations (3.7) and (3.8).
Proof. See [37]. 
We now present a simulated example for showing the performance of the
adaptive controller.
Simulation experiment 3.2. In this experiment, the elements of B were
given as b(11) = 4, b(12) = 2, b(21) = 2, b(22) = 1 (detB = 0).The interval
82 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

Figure 3.11 Configuration of adaptive stabilization system

estimates of these elements were chosen as follows: 1 ≤ b(11) ≤ 5,

0 ≤ b(12) ≤ 2, 0 ≤ b(21) ≤ 2, 1 ≤ b(22) ≤ 2. By the formulas (3.42) and
(1) (2)
(3.44), it was found: β (1)
min
= −1, β (2)
min
= −1, β max = 7, β max = 4, β min = −1,
β max = 7. It turned out that |β min | < | β max |. Therefore it is required that
δ0 > 1 to satisfy the inequalities (3.43). Namely, δ0 = 1.1 was put. From the
(11) (12) (21) (22)
conditions b0 ∈ [1, 5], b0 ∈ [0, 2], b0 ∈ [0, 2], b0 ∈ [1, 2]
(11)
the following initial estimates of the elemetns of Bn were taken: b0 = 1,
(21) (12) (22) (11) (12) (21)
b0 = 0, b0 = 1, b0 = 1.9. Then b̃0 = 2.1, b̃0 = 1, b̃0 =
(22)
0, b̃0 = 3.
(i)
In this simulation experiment, the sequences {vn }(i = 1, 2) were
generated as i.i.d. random variables belonging to [−1, 1]. It was put: y 0 =
[1, 3]T .
 3.5 Robustly-Adaptive Control of Square Systems (Case 1) 83

The performance of the adaptive estimation algorithm is shown in

Figure 3.12.
Figure 3.13 shows simulation results illustrating a successful behavior of
the robustly-adaptive control system when this experiment was conducted.

Figure 3.12 Estimated parameters of the fictitious plant (solid line) and the true plant
(dashed line)

Figure 3.13 Performance of the robustly-adaptive controller in Simulation experiment 3.2:

(a) Euclidean norm of input control vector; (b) Euclidean norms of output vector
84 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

3.6 Robustly-Adaptive Control of Square Systems

(Case 2)
(i)
In this case called Case 2 we assume that the bounds ε(i) on disturbances vn
are unknown a priori. Therefore, they need to be adaptively estimated.
Following to [38], the algorithm for the adaptive estimation of the
fictitious parameters is chosen as
∗(i) (i)
f (ẽn , εn−1 )
∇un sign ẽ∗(i)
(i)
n = b̃n−1 − γn
b̃(i) (i)
n , i = 1, . . . , r (3.53)
1 + ||∇un ||2

where f (e, ε) is the dead-zone function of the form


0, if|e| ≤ ε,
f (e, ε) =
|e| − ε if|e| > ε
∗(i)
which depends on the scalar variable ẽn given by (3.51) and also on the past
(i)
estimate εn−1 of unknown ε(i) .
As in [38] the adaptive estimation of the bounds ε(i) is produced via the
recursive algorithm
∗(i) (i)
(i) f (ẽn , εn−1 )
ε(i) (i)
n = εn−1 + γn , i = 1, . . . , r (3.54)
1 + | |∇un | |22
The following results concerning the asymptotic properties of the robustly-
adaptive control system for Case 2 are valid.
Theorem 3.4. Let the conditions of Theorem 3.3 be satisfied. Then the
adaptive controller described in Equations (3.47), (3.48), (3.53) and (3.54)
has the properties:
(a) the estimates caused by Equations (3.53) and (3.54) converge;
(b) the ultimate boundedness in the forms (3.7) and (3.8) are achieved.

Proof. See [38]. 

Simulation experiment 3.3. The conditions of experiment were taken as in
(1) (2)
previous experiment except the bounds on the disturbance vn and vn given
(1) (2)
by −1 ≤ vn ≤ 1 and −0.5 ≤ vn ≤ 0.5.
Results of this simulation experiment are presented in Figures 3.14–3.17.
 3.6 Robustly-Adaptive Control of Square Systems(Case 2) 85

(11) (12) (21) (22)

Figure 3.14 Estimates b̃n , b̃n , b̃n and b̃n of Simulation experiment 3.3

(1) (2)
Figure 3.15 Estimates εn /2 and εn /2 of Simulation experiment 3.3

It is seen from Figures 3.14 and 3.15, that the estimates of unknown
system parameters and bounds on disturbances converges as predicted by part
(a) of Theorem 3.4. We can also observe that the control inputs and outputs
remain bounded.
86 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

(1) (2)
Figure 3.16 Control inputs un and un of Simulation experiment 3.3

(1) (2)
Figure 3.17 Outputs yn and yn of Simulation experiment 3.3

3.7 Robustly-Adaptive Control of Nonsquare Systems

Let B be a nonsquare m × r matrix of the form (3.2) with unknown rank sat-
isfying (3.4). Define the so-called submatrices B[i1 [k], . . . , ir [k]|1, . . . , r]
∈ Rr×r [43, part I, subsect. 2.2] whose rows represent the rows of B with the
numbers i1 [k], . . . , ir [k] (1 ≤ i1
[k] < · · · < ir [k] ≤ m). The quantity
m
of these matrices is equal to N = . Denoting by B[k] the submatrix
r
which corresponds to a kth subset {i1 [k], . . . , ir [k]}, write the equations of
some k plants as:
yn [k] = B[k]un−1 + vn [k], k = 1, . . . , N, (3.55)
(i [k]) (i [k]) T (i [k])
where yn [k] = [yn 1 , . . . , yn r ] ∈ Rr and vn [k] = [vn 1 ,...,
(i [k])
vn r ] T ∈ R r .
 3.7 Robustly-Adaptive Control of Nonsquare Systems 87

In accordance with the approach proposed in the previous section, pass

from Equation (3.55) to the equations of the fictitious plants described by

ỹn [k] = B̃[k]un−1 + vn [k], k = 1, . . . , N (3.56)

with the same un−1 and vn−1 [k]. In these equations, ỹn [k] denotes the r-
dimensional output vector related to the kth fictitious plant whose gain matrix
B̃[k] is defined as follows:

B̃[k] = B[k] + δ0 [k]Ir , (3.57)

where δ0 [k] is a fixed quantity depending on k. This quantity is calculated for

each k = 1, . . . , N using the technique described in the previous section.
Namely, taking into account the constraints (3.5), δ0 [k] can always be found
to ensure
det B̃[k] = 0 ∀k = 1, . . . , N. (3.58)
It follows from Equations (3.55) to (3.57) that

ỹn [k] = yn [k] + δ0 [k]un−1 . (3.59)

This expression shows that although as B̃[k] as B[k] remain unknown, how-
ever, the components of all N the vectors ỹn [k] can indirectly be “measured”
after measuring the components of yn and un−1 , and it is essential.
If the conditions (3.58) are satisfied, then the problem of the adaptive
stabilization of the true plant (3.1) can be reduced to the problem of simul-
taneous adaptive stabilization of all N fictitious plants (3.56) with unknown
but nonsingular r × r gain matrices B̃[k] (k = 1, . . . , N ) via forming
at each nth time instant a set of N different “potentially” possible controls
un [1], . . . , un [N ] and selecting one of them in accordance with certain
choice rule [27] given below.
Following to [39], the adaptive control law to be applicable to any
fictitious plant is designed in the form

un [k] = un−1 + B̃n−1 [k] ẽn [k], k = 1, . . . , N, (3.60)

where ẽn [k] = y 0 [k] − ỹn [k] with y 0 [k] = [y 0(i1 [k]) , . . . , y 0(ir [k]) ]T defines
the output error vector related to the kth fictitious plant at the nth time instant,
and B̃n [k] ∈ Rr×r is the current estimate of unknown r × r matrix B̃[k] at
the same time instant satisfying

det B̃ n [k] = 0 ∀k = 1, . . . , N. (3.61)

88 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

As the adaptation algorithms, the standard recursive procedures for the

adaptive identification of each kth fictitious plant (3.55) described by
⎧
∗(i)
⎪
⎪
(i)
b̃n−1 [k] if |ẽn [k]| ≤ ε0i ,
⎪
⎪
⎨
b̃(i) [k] = (i) (i) ∗(i) ∗(i)
n
⎪ b̃n−1 [k] + γn ẽn [k]−ε i sign ẽn [k]
2 ∇un−1 otherwise,
⎪
⎪ ||∇u n−1 || 2
⎪
⎩
i = 1, . . . , r, k = 1, . . . , N
(3.62)
(i)
are proposed. In these algorithms, b̃n [k] denotes the r-dimensional estimate
vector obtained by transposing the ith row of B̃n [k], and

ẽ∗(i)
(i) (i)T
n [k] = ỹn [k] − ỹn−1 [k] − b̃n−1 [k]∇un−1
(i)
(3.63)
represents the scalar variable making sense of the ith component of ẽ∗n [k] ∈
Rr that is the identification error vector related to the kth fictitious plant.
(i)
The coefficients γn s are chosen from the ranges [γ (i) , γ (i) ] to ensure the
requirement (3.61).
Next, add the adaptation algorithms described in the formulas (3.62)
together with (3.63) by an algorithm for estimating unknown B defined as
follows:
⎧
∗(i)
⎪ (i)
⎨ bn−1 if |en | ≤ ε0i ,
b(i)
n =
⎪ ∗(i) ∗(i)
⎩ b(i) + γn(i) en −εi sign e2n ∇un−1 otherwise,
n−1 ||∇un−1 ||2
i= 1, . . . , m,
(3.64)
(i)T
where bn represents the ith row of the estimate matrix Bn , and

e∗(i)
(i) (i)T
n = yn(i) − yn−1 − bn−1 ∇un−1 (3.65)
is the ith component of the identification error vector e∗n = yn − yn−1 −
Bn−1 ∇un−1 (εi and ε0i are given by the conditions (3.50)).
The estimation procedure defined in the algorithm (3.64) together with
Equation (3.65) makes it possible to estimate the m predicted output errors
→
− (i)
e n+1 [k] (i = 1, . . . , m) for the each ith output of true plant (3.1) at any n
using the formula

|−
→ (i)
e n+1 [k]| = |y 0(i) − b(i)T (i)
n un [k]| + ε , i = 1, . . . , m. (3.66)
The synthesis of the adaptive controller is finished by the choice of
the control un from the set {un [1], . . . , un [N ]} with un [k] given by
 3.7 Robustly-Adaptive Control of Nonsquare Systems 89

Equation (3.60). This choice is implemented by the rule giving the minimum
of the 1-norm of −
→e n+1 [k] = [−
→
e n+1 [k], . . . , −
→
(1) (m)
e n+1 [k]]T according to

!
m
|−
→ (i)
un = arg min e n+1 [k]|, (3.67)
un [k]
i=1

where −→ (i)
e n+1 [k]s are specified by Equation (3.66).
The asymptotic properties of the adaptive controller described in this
section are given in theorem below (the main result).
Theorem 3.5. Consider the feedback control system containing the plant (3.1)
in which r < m, and the adaptive controller defined in Equations (3.62),
(3.67) together with Equations (3.59), (3.66) and (3.61). Using the constraints
(3.5), determine δ0 [1], . . . , δ0 [N ] to satisfy the requirement (3.58). Let
assumptions A1–A3 be valid. Then, this controller applied to the plant (3.1)
guarantees that the control objectives (3.7) and (3.8) will be achieved.
Proof. See [39]. 
Note that Theorem 3.5 does not guarantee that the ultimate error
limn→∞ sup ||en || will become as in the nonadaptive case when there
is no parametric uncertainty and the pseudoinverse model-based controller
proposed in [21] can by applied.
Simulation experiment 3.4. A simulation experiment was conducted to illus-
trate the performance of the proposed adaptive control in the case when
r = 2, m = 3. As the gain matrix,
⎛ ⎞
4 2
B=⎝ 2 1 ⎠
3 1.5

of not full rank (rank B= 1) was taken. Since N = 3, it produces the

following three submatrices:
     
4 2 4 2 2 1
B[1] = , B[2] = and B[3] =
2 1 3 1.5 3 1.5
(1) (2) (1) (3)
Further, the three vectors yn [1] = [yn , yn ]T , yn [2] = [yn , yn ]T and
(2) (3)
yn [3] = [yn , yn ]T was introduced to describe the plants (3.55) having the
gain matrices B[1], B[2] and B[3], respectively.
90 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

Figure 3.18 Variables describing the adaptive estimation processes

Figure 3.19 The behaviour of the control system: (a) the current number k of control un [k]
chosen from {un [1], un [2], un [3]} at given n; (b) the 1-norm of control vector; (c) the 1-norm
of output vector in adaptive case (solid line) and in nonadaptive optimal case (dashed line)
 References 91

The quantities δ0 [1] = 1.1, δ0 [2] = 1.2 and δ0 [3] = 1.3 were taken
to satisfy the conditions (3.58) guaranteeing B̃[k] to be nonsingular were
derived from Equation (3.5). The initial B̃0 [1], B̃0 [2] and B̃0 [3] were chosen
as B̃0 [k] = B0 [k] + δ0 [k]Ir with the initial elements of B0 [k] which were
(ij)
selected from B inside the corresponding ranges [b(ij) , b ] specified as
follows: b(11) ∈ [1, 5], b(12) ∈ [0, 2], b(21) ∈ [0, 2], b(22) ∈ [1, 2],
(11) (12) (21)
b(31) ∈ [1, 4], b(32) ∈ [0, 5]. Namely, we set b0 = 1, b0 = 1 b0 = 0,
(22) (31) (32)
b0 = 1.9, b0 = 2, b0 = 2.1. The desired output vector was given as
y 0 = [1, 3, 7]T .
The performance of the simulated adaptive control system with the dis-
(i) (i) (i)
turbance sequences {vn } = v0 , v1 , . . . generated as some i.i.d. random
(1) (2)
variables taken from −0.1 ≤ vn ≤ 0.1, −0.2 ≤ vn ≤ 0.2, −0.08 ≤
(3)
vn ≤ 0.08 is presented in Figures 3.3 and 3.19.
Figures 3.19(a)–(c) demonstrate that the performance of the proposed
adaptive controller applied to the static multivariable plant having some
nonsquare gain matrix with not full rank is successful enough.

3.8 Conclusion
This chapter shed a light on the problem regarding the robust control of the
discrete-time multivariable memoryless systems with the noninvertible gain
matrices of the not full rank. It has been established that if certain condi-
tions have been satisfied then the nonadaptive pseudoinverse model-based
controller can be robust against plant uncertainty.
Again, the chapter answered the second question: Can robustly-adaptive
controller using as adjustable inverse model principle cope with the interval
parametric uncertainty of the system to be controlled? Namely, it has been
shown that such a possibility exists because these controllers are able to
guarantee the ultimate boundedness of output and control input signals.
However, this important feature is achieved via an “overparametrization” of
the uncertain systems. Nevertheless, the simulation experiments showed their
efficiency.

References
[1] V.M. Kuntsevich, et al (Eds). Control Systems: Theory and Applications.
Series in Automation, Control and Robotics, River Publishers, Gistrup,
Delft, 2018.
92 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

[2] Y.P. Kondratenko, V.M. Kuntsevich, A.A. Chikrii, V.F. Gubarev, (Eds).
Advanced Control Systems: Theory and Applications. Series in Automa-
tion, Control and Robotics, River Publishers, Gistrup, 2021.
[3] S. Skogestad and I. Postlethwaite, Multivariable Feedback Control,
Wiley, Chichester, 1996.
[4] P. Albertos and A. Sala, Multivariable Control Systems: an Engineering
Approach, Springer, London, 2006.
[5] T. Glad and L. Ljung, Control Theory: Multivariable and Nonlinear
Methods, Taylor & Francis, New York, 2000.
[6] L. Tan, A Generalized Framework of Linear Multivariable Control,
Elsevier, Oxford, 2017.
[7] V.F. Gubarev, M.D. Mishchenko, and B.M. Snizhko, ‘Model predic-
tive control for discrete MIMO linear systems’, In: Kondratenko, Y.P.,
Chikrii, A.A., Gubarev, V.F., Kacprzyk, J. (Eds.). Advanced Control
Techniques in Complex Engineering Systems: Theory and Applications.
Dedicated to Prof. V.M. Kuntsevich. Studies in Systems, Decision and
Control, pp. 63-81, vol. 203, Springer Nature Switzerland AG, Cham,
2019. DOI: https://doi.org/10.1007/978-3-030-21927-7_10
[8] E. Davison, ‘The output control of linear time-invariant multivariable
systems with unmeasurable arbitrary disturbances’, IEEE Trans. Autom.
Contr., pp. 621–631, no. 5, vol. AC-17, Oct. 1972.
[9] B. Francis and W. Wonham, ‘The internal model principle of control
theory’, Automatica, pp. 457–465, no. 5, vol. 12, Sep. 1976.
[10] R. Brockett, ‘The invertibility of dynamic systems with application to
control’, Ph. D. Dissertation, Case Inst. of Technology, Cleveland, Ohio,
1963.
[11] M.K. Sain and J.L. Massey, ‘Invertibility of linear time-invariant
dynamical systems’, IEEE Trans. Autom. Contr, pp. 141–149, no. 2,
vol. AC-14, Apr. 1969.
[12] L.M. Silverman, ‘Inversion of multivariable linear systems’, IEEE
Trans. Autom. Contr., pp. 270–276, no. 3, vol. AC-14, Jun. 1969.
[13] H. Seraji, ‘Minimal inversion, command tracking and disturbance
decoupling in multivariable systems’, Int. J. Control, pp. 2093–2191,
no. 6, vol. 49, Jun. 1988.
[14] G. Marro, D. Prattichizzo, and E. Zattoni, ‘Convolution profiles for
right-inversion of multivariable non-minimum phase discrete-time sys-
tems’, Automatica, pp. 1695–1703, no. 10, vol. 38, Oct. 2002.
 References 93

[15] L.M. Lyubchyk, ‘Disturbance rejection in linear discrete Multivariable

systems: inverse model approach’, In: Proc. 18th IFAC World Congress,
Milano, Italy, 2011, pp. 7921–7926.
[16] L.M. Lyubchyk, ‘Inverse Model Approach to Disturbance Rejection
Problem’, In: Kondratenko, Y.P., Kuntsevich, V.M., Chikrii, A.A.,
Gubarev, V.F. (Eds), Advanced Control Systems:Theory and Applica-
tions, River Publishers, Gistrup, pp. 129–166, 2021.
[17] G.E. Pukhov and K.D. Zhuk, Synthesis of Interconnected Control
Systems via Inverse Operator Method, Nauk. dumka, Kiev, 1966 (in
Russian).
[18] T. Lee, G. Adams, and W. Gaines, Computer Process Control: Modeling
and Optimization, Wiley, New York, 1968.
[19] V. Lovass-Nagy, J.R. Miller, and L.D. Powers, ‘On the application of
matrix generalized inversion to the construction of inverse systems’, Int.
J. Control, pp. 733–739, no. 5, vol. 24, Apr. 1976.
[20] L.S. Zhiteckii, V.N. Azarskov, K.Yu. Solovchuk and O.A. Sushchenko,
‘Discrete-time robust steady-state control of nonlinear multivariable
systems: a unified approach’, In: Proc. 19th IFAC World Congress, Cape
Town, South Africa, 2014, pp. 8140–8145.
[21] L.S. Zhiteckii and K.Yu. Solovchuk, ‘Pseudoinversion in the problems
of robust stabilizing multivariable discrete-time control systems of linear
and nonlinear static objects under bounded disturbances’, Journal of
Automation and Information Sciences, pp. 35–48, no. 5, vol. 49, Jul.
2017.
[22] S.P. Bhattacharyya, H. Chapellat, and L.H. Keel, Robust Control: the
Parametric Approach. Prentice Hall, Upper Saddle River, NJ, 1995.
[23] M. Green and D.J.N. Limebeer, Linear Robust Control. Prentice Hall,
Englewood Cliffs, NJ, 1995
[24] R. Sanchess-Pena and M. Sznaier, Robust Systems: Theory and Appli-
cation. Willey, New York, 1998.
[25] B.T. Polyak and P.S. Shcherbakov, Robust Stability and Control, Nauka,
Moscow, 2002 (in Russian).
[26] V.M. Kuntsevich, Control under Uncertainty: Guaranteed Results in
Control and Identification Problems, Nauk. dumka, Kiev, 2006 (in
Russian).
[27] V.F. Sokolov, Robust Control with Bounded Disturbances, Komi Sci-
entific Center, Ural Branch of the Russian Academy of Sciences,
Syktyvkar, 2011 (in Russian).
94 Robust and Robustly-Adaptive Control of Some Noninvertible Memoryless

[28] V.M. Kuntsevich, ‘Estimation of impact of bounded perturbations on

nonlinear discrete systems’, In: Kuntsevich, V.M., Gubarev, V.F., Kon-
dratenko, Y.P., Lebedev, D.V., Lysenko, V.P. (Eds), Control Systems:
Theory and Applications. Series in Automation, Control and Robotics,
River Publishers, Gistrup, Delft, pp. 3-15, 2018.
[29] B.I. Kuznetsov, T.B. Nikitina, and I.V. Bovdui, ‘Multiobjective synthesis
of two degree of freedom nonlinear robust control by discrete continu-
ous plant. Tekhnichna elektrodynamika’, Institute of Electrodynamics
National Academy of Science of Ukraine, pp. 10–14, no. 5, Sep. 2020.
[30] V.N. Fomin, A.L. Fradkov, and V.A. Yakubovich, Adaptive Control of
Dynamic Plants, Nauka, Moscow, 1981 (in Russian).
[31] G.C. Goodwin and K.S. Sin, Adaptive Filtering, Prediction and Control,
Prentice-Hall, Engewood Cliffs, 1984.
[32] G. Tao. Adaptive Control Design and Analysis. John Wiley and Sons,
New York, 2003.
[33] K.S. Narendra and A.M. Annaswamy, Stable Adaptive Systems, Dover
Publications, New York, 2012.
[34] P. Ioannou and J. Sun, Robust Adaptive Control, Dover Publications,
New York, 2013.
[35] K.J. Åström and B. Wittenmark, Adaptive Control: 2nd Edition, Dover
Publications, New York, 2014.
[36] L.S. Zhiteckii and K.Yu. Solovchuk, ‘Robust Adaptive Controls for
a Class of Nonsquare Memoryless Systems’, In: Kondratenko, Y.P.,
Kuntsevich, V.M., Chikrii, A.A., Gubarev, V.F. (Eds), Advanced Con-
trol Systems:Theory and Applications, River Publishers, Gistrup, pp.
203–226, 2021.
[37] V.N. Azarskov, L.S. Zhiteckii, and K.Yu. Solovchuk, ‘Parametric identi-
fication of the inter-connected static closed-loop system: a special case’,
In: Proc. 12th All-Russian Control Problems Council (VSPU-2014),
IPU, Moscow, 2014, pp. 2764-2776.
[38] V.N. Azarskov, L.S. Zhiteckii, and K.Yu. Solovchuk, ‘Identification
approach to the problem of robust control for multi-connected static
plant with nonstochastic uncertainties’, In: Proc. 10th Internation Con-
ference System identification and Control Problems (SICPRO’15), IPU,
Moscow, 2015, pp. 520-538.
[39] L.S. Zhiteckii, V.N. Azarskov, and K.Yu. Solovchuk, ‘Adaptive robust
control of inter-connected static plants with nonsquare gain matrixes’,
In: Proc. 13th All-Russian Control Problems Council (VSPU-2019),
IPU, Moscow, 2019, pp. 713–718.
 References 95

[40] L.S. Zhiteckii and K.Yu. Solovchuk, ‘Robustly-adaptive control of lin-

ear interconnected memoryless systems with nonsquare gain matrices’,
In: Proc. Int. Conf. “Automatics–2020”, Kyiv, Ukraine, pp. 123–124,
2020.
[41] A. Albert, Regression and the Moore-Penrose Pseudoinverse, Academic
Press, New York, 1972.
[42] V.V. Voevodin and Yu.A. Kuznetsov, Matrices and Calculations, Nauka,
Moscow, 1984 (in Russian).
[43] M. Marcus and H. Minc, A Survey of Matrix Theory and Matrix
Inequalities, Aliyn and Bacon, Boston, 1964.
 4
Nonlinear Integral Inequalities
and Differential Games
of Avoiding Encounter

L.P. Yugay

Dr.Sc.(Math), Professor,
Department of Natural Sciences, Uzbek State University of Physical Culture
and Sport, Chirchik City, Republic of Uzbekistan
Tel: +998-90-187-1478(m)
E-mail: [email protected]

Abstract
In the paper, a survey of the application of integral inequalities in various the-
oretical branches of mathematics and their applied fields is represented. New
integral inequalities of the Hölder type are introduced, and corresponding
estimates are proved. Also, the possibilities of applications of demonstrated
results in nonlinear differential games of avoiding encounters are indicated.

Keywords: integral inequality, estimation, trajectory, differential game, eva-

sion, avoiding, trajectory.

4.1 Introduction
4.1.1 On the First Integral Inequalities
Integral inequalities are widely used in the theories of differential equa-
tions, nonlinear oscillations, stability, as well as in mechanics and averaging
problems. With their help, the unique theorems of solutions of differential
equations are proved, and various approximate estimates are obtained for
solutions of differential equations that cannot be analytically solved. Let us
formulate the first variance of Grönwall–Bellman type inequality.

97
98 Nonlinear Integral Inequalities and Differential Games

Theorem 1 (Inequality of Grönwall–Bellman)

Let u (t) and v (t) are nonnegative and continuous functions on the length
[a, b] and
 b
v (t) ≤ C + v (s) u (s) ds, a ≤ t ≤ b, C ≥ 0. (4.1)
a

Then  b
v (t) ≤ Cexp u (s) ds. (4.2)
a
The inequality (4.2) was proved by T.H. Grönwall in [1] (1919) and extended
by R. Bellman in [2] (1943). We must note that Grönwall–Bellman inequality
(4.2) was the beginning point for studying many integral inequalities and their
applications. Particularly, in [3], P. Hartman writes that the result (4.2) has
as its sours the work of G. Peano [4]. Among the publications on nonlinear
Grönwall–Bellman type inequalities, the first papers were published by Bihari
[5] (1956) and A. Perov [6] (1957). Therefore integral inequalities of this
type (linear and nonlinear) are usually associated with Grönwall, Bellman,
Bihari, and others. The other names to be mentioned in connection with the
further development of the theory of integral inequalities are R.P. Agarval,
S.A. Chaplygin, N.B. Azbelev, D.D. Bainov, P.R. Beesack, S.G. Deo, U.D.
Dhongade, V. Lakshmikantham, S. Leela, B.G. Pachpatte, J. Popenda, P.S.
Simeonov, Z.B. Tsalyuk, D. Willett, J.S.W. Wong, E.C. Young. Wong, E.C.
Young. Integral inequalities are necessary for studying various equations (dif-
ferential, difference, discrete, etc.). This is why the intensive investigations of
different integral inequalities and hundreds of publications on them appear. A
part of the results can be found in the monographs of E.F. Beckenbach and R.
Bellman (1965), V. Lakshmikantham and S. Leela (1969), W. Walter (1970),
P.R. Beesack (1975), A.N. Filatov and L.V. Sharova (1976), A.A. Martynyuk
and R. Gutowski (1979), R. Rabczuk (1979), D.Ya. Mamedov, S. Ashirov, S.
Atdaev (1980), and J. Schroder (1980).

4.1.2 Development of the Theory of Integral Inequalities (Brief

Survey)
The theory of integral inequalities development can be conditionally divided
into three stages. At the first stage, linear integral inequalities were developed
and were widely applied in differential equations research. One of the first
publications in this direction [1–3] began research on linear integral inequal-
ities (4.1-4.2) of the Grönwall–Bellman type and their generalizations to
 4.1 Introduction 99

various linear and nonlinear differential equations classes. During this period,
integral inequalities constituted the central part of the proofs of existence and
uniqueness theorems for solutions of nonlinear differential equations. They
were also used in evaluating their solutions (one-sided and two-sided) [5–13].
Using integral inequalities, approximate estimates for solutions of differential
equations that do not have analytical solutions are obtained.
Primary applications of integral inequalities are outlined in the actual
work [10].
The next stage in the development of research of integral inequalities
dates back to the 1970–80s. Integral inequalities began to be used for the
solution of problems in the theory of nonlinear oscillations and the theory of
stability, as well as in mechanics, averaging problems, etc. A detailed review
of nonlinear integral inequalities can be found in [10], and applications in
nonlinear oscillations and mechanics are presented in [11–13].
As we can see from the listed sections, there are no applications of integral
inequalities in such important applied areas as the theory of optimal control
and its extension of the theory of conflict-controlled systems (differential
games). Perhaps this is due to the thematic interests of the authors because
integral inequalities are widely used in the theory of nonlinear optimal and
automatic control problems [14–18] and the theory of differential escape
games [19–59].
The theory of integral inequalities is intensively developed (third stage).
It covers new areas of applications, in particular, the theory of optimal
(automatic) control [14–18] and the theory of nonlinear differential games
[19–59].
In the theories of optimal and automatic control, as well as in differential
games, integral inequalities are applied to:
• identification (estimation) of trajectories of nonlinear controlled and
conflict controlled systems [14–30, 52, 58];
• obtaining lower estimates for the distances from the current position of
the trajectory to the terminal (target) set [31–42, 53–56];
• estimates of admissible controls that satisfy “energy” (integral) con-
straints and integral inequalities [39, 44–47, 58, 59].
Recently, there has been an active extension of new applications of
integral inequalities in differential games [48]. Thus, in [48–50], the authors,
for the first time, consider new classes of controls for the game partic-
ipants (the pursuer and the evader) based on integral inequalities of the
Grönwall–Bellman type. These controls generalize well-known controls
100 Nonlinear Integral Inequalities and Differential Games

satisfying integral constraints [39] and allow to consider differential games

with geometrical and integral constraints from a unified position, In the
past geometrical and integral constraints on controls were considered in
differential games separately.
Note that the newly introduced classes of controls [48–50] are mainly
used in pursuit problems of differential games only.

4.2 Investigation of Nonlinear Hölder Type Integral

Inequality
4.2.1 Main Results
Lemma 4.1. Let the function

z (ξ) = ξ − γ1 ξ α1 − γ2 ξ α2 − · · · − γn ξ αn − γn+1 ,

where ξ ≥ 0, γi > 0 are constants,i = 1, 2, . . . n + 1; 0 < αj < 1, j =

1, 2, . . . , n.
Then the function z (ξ)has a unique minimum point and a unique root
ξ0 > 0and is also strictly convex for all ξ ≥ 0.. In addition, for all 0 ≤ ξ ≤ ξ0 ,
the inequality

z (ξ) = ξ − γ1 ξ α1 − γ2 ξ α2 − · · · − γn ξ αn − γn+1 ≤ 0 (4.3)

is hold.

Proof. Let’s first investigate the extremum of z (ξ). To do this, we calculate

the derivatives of the function, which are equal:
!
n !
n
z (ξ) = 1 − γi αi ξ αi −1 , z (ξ) = γi αi (1 − αi ) ξ αi −1 . (4.4)
i=1 i=1

It’s easy to see that

z (0) = −γn+1 < 0 (4.5)
and
!
n
lim z (ξ) = lim ξ(1 − γi ξ αi −1 − ξ −1 γn+1 ) = +∞. (4.6)
ξ→+∞ ξ→+∞
i=1

Relations (4.4)–(4.6) show that, due to continuity z (ξ), there exists a

positive root ξ0 > 0 of the function z (ξ), that is z (ξ0 ) = 0.
 4.2 Investigation of Nonlinear Hölder Type Integral Inequality 101

Next, let’s find out the behavior of the derivative. We have, taking into
account 0 < αj < 1, j = 1, 2, . . . , n, that
!
n
lim z (ξ) = lim (1 − γi αi ξ αi −1 ) = 1,
ξ→+∞ ξ→+∞
i=1

!
n
1

lim z (ξ) = lim (1 − γi αi ) = −∞.
ξ→+0 ξ→+0 ξ 1−αi
i=1
From this and the continuity of the function z (ξ), we obtain that there is a
point, such that z (ξ∗ ) = 0.
Further investigation of the critical point ξ ∗ shows that at this
point z ( ξ ∗ ) > 0, therefore, ξ ∗ is a point of a local minimum. Further,
since z (ξ) > 0 for all ξ > 0 (including the point ξ = ξ ∗ ) then z (ξ) it is
a strictly convex function for all ξ ≥ 0, and such functions, the extremum
point ξ∗ > 0 is the only point of the global minimum of the function z (ξ)
ξ ≥ 0. in the domain ξ ≥ 0. It is easy to verify that ξ 0 > ξ∗ > 0.
Let 0 ≤ ξ ≤ ξ 0 , then from the property of the root ξ 0 we have, which
proves (4.3). Lemma 1 is proved.

Theorem 4.1 (Main result). Let the function u (t) be continuous for t0 ≤
t ≤ t0 + 1, t0 ≥ 0, and satisfies the Hölder type integral inequality (4.7):
 t !
n

0 ≤ u (t) ≤ f (t) + a (s) + b (s) u (s) + ci (s)uαi (s) ds (4.7)
t0 i=1

where all functions f (t) , a (t) , b (t) ci (t) are given, nonnegative, inte-
grable, and are not equal identically to zero on the interval [t0 , t0 +1],
0 < αi < 1, i = 1, 2, . . . n.
Then there is a constant h0 ∈ (t0 ; t0 + 1], such that the function u(t) for
all t ∈ [t0 ; t0 + h0 ] satisfies the inequality
t !
n
u (t) ≤ f (t) + a (s) ds+δξ0 + ξ0αi gi (t) , (4.8)
t0 i=1

where ξ0 is the positive root of the equation

!
n
ξ− di ξ αi − dn+1 = 0, (4.9)
i=1
102 Nonlinear Integral Inequalities and Differential Games
⎛ ⎞1−αi
t 1
gi (t) = ⎝ ci 1−αi
(s) ds ⎠ ,
t0

and constants δ, h0 , b1 , di , i = 1, 2, . . . , n + 1, are positive and calculated

effectively by the functions and the parameters αi ∈ (0; 1) given in (4.7).

Proof. Without loss of generality, we can assume in Theorem 1 t0 = 0.

Let’s positive. From (4.7), by using Hölder’s integral inequality, we obtain
that for all 0 ≤ t ≤ h ≤ 1 the inequalities
t t t 1!
n
2
u (t) ≤ f (t) + a (s) ds + δ u (s) ds + ci (s)uαi (s) ds ≤
0 0 0 i=1

⎛ ⎞ αi ⎛ ⎞1−αi
t h !n h t
1 1
f (t)+ a (s) ds+δ u (s) ds+ ⎝ (uαi (s)) αi
ds)⎠ ⎝ ci 1−αi
(s) ds)⎠ ≤
0 0 i=1 0 0
⎛ ⎞ αi ⎛ ⎞1−αi
t h !
n h t
1
f (t) + a (s) ds + δ u (s) ds + ⎝ u(s) ds)⎠ ⎝ ci 1−αi
(s) ds)⎠
0 0 i=1 0 0

are fulfilled, which immediately give us the inequality

⎛ h ⎞ ai
t h !n 
u (t) ≤ f (t) + a (s) ds + d u (s) ds + ⎝ u (s) ds)⎠ gi (t) .
0 0 i=1 0
(4.10)
Further, we will estimate the integral
h
ξ= u (s) ds)
0

in (4.10). To do this, we integrate (10) once again with 0 ≤ t ≤ h ≤ 1, then

we have:
h h t h
u (t) dt ≤ [f (t) + a (s) ds]dt + δh u (s) ds
0 0 0 0
⎛ ⎞αi
!
n h h
+ ⎝ u (s) ds)⎠ gi (t) dt. (4.11)
i=1 0 0
 4.2 Investigation of Nonlinear Hölder Type Integral Inequality 103

Let introduce the following notations:

1 1 1
γi = gi (t) dt, i = 1, 2, . . . n, γ n+1 = [f (t) + a (s) ds]dt,
0 0 0
then from (4.11), one has
!
n
ξ≤ δhξ+ γ i ξ αi + γ n+1 ,
i=1
or
!
n
(1 − δh) ξ ≤ γ i ξ αi + γ n+1 . (4.12)
i=1
Let us define the following constants:
1 1 1 1
h0 = min{1; }, γi = γi , γn+1 = γ .
2 δ 1 − δh0 1 − δh0 n+1
Then for each fixed, h ∈ [0; h0 ] the inequality 1 − δh > 0 is fulfilled, and
(11) can be written in the form
!
n
ξ− γi ξ αi − γn+1 ≤ 0. (4.13)
i=1
Note that in (13) ξ = ξ(h), where h ∈ [0; h0 ] , γi and γn+1 are positive
constants, not depending on.
If we put ξ (h0 ) = x, then from (4.13) we obtain for h = h0 the
inequality
!n
x− γi xαi − γn+1 ≤ 0. (4.14)
i=1
Obviously, in (4.14) x is a variable depending on the function u (t) , t ∈
[0; h0 ] .
Applying Lemma 1 to (4.14),  we have that there is a unique root ξ0 >
0 of the function z (x) = x − ni=1 γi xαi − γn+1 , such that inequality
(4.14) holds for all 0 ≤ x ≤ ξ0 , this implies for all h ∈ [0; h0 ] inequalities
h h0
u (s) ds) ≤ u (s) ds = x ≤ ξ0 . (4.15)
0 0
Applying (4.15) to (4.10), we get the required inequality (4.8). Theorem 1 is
proved.
104 Nonlinear Integral Inequalities and Differential Games

4.2.2 Applications of Integral Inequalities in Differential Games

of Avoiding an Encounter
Let a vector differential equation describe conflict controlled process (differ-
ential game)
ż = f (z, u, v) , (4.16)
where z ∈ Rn , u ∈ P ⊂ Rp , v ∈ Q ⊂ Rq , P and Q are non-empty compact
sets, containing starting points 0p ∈ Rp and 0q ∈ Rq , a function f (z, u, v)
is continuous over X = Rn × P × Q and satisfies for any (z, u, v) ∈ X
to inequalities  
2
< z, f (z, u, v) >≤ C 1 + |z| ,

|f ( z 2 , u, v) − f ( z 1 , u, v)| ≤ Lα | z 2 − z 1 |α , (4.17)
where z 1 ∈ Rn , z 2 ∈ Rn , 0 < α ≤ 1, C ≥ 0 and Lα ≥ 0 are constants.
Terminal set M is a linear subspace of Rn .
Parameters u and v in (4.16) are chosen by opposing sides (players) as
measurable functions u = u (t) ∈ P, v = v (t) ∈ Q, t ≥ 0. Player,
choosing v = v (t) ∈ Q (evader), sets itself a task for any acceptable
behavior u (t) ∈ P to avoid encountering the corresponding trajectory z(t)
of equation (4.16), starting from any point z0 ∈ Rn \M (z0 = z(0)),
from M for any t ≥ 0. Such problem is called global avoiding encounter
problem(global evasion problem), which was formulated by L.S. Pontrjagin
and E.F. Mis̆c̆enko in [31]. At every moment t ≥ 0 the evader by forming
value of v (t) ∈ Q may use values u (s) ∈ P and z(s) with s ≤ t, but he
does not know its future values at s > t.
In addition, according to the conditions of the avoiding problem, the
evading player has a particular advantage over the pursuer for the successful
solution of the collision avoidance problem. The benefits are set in differ-
ent ways (geometric, integral, mixed, etc.) and are usually included in a
conflict-controlled process’s dynamics (4.16).
Using its advantages in dynamics and information, the evader firstly
constructs a local special control function v c (t) ∈ Q, which ensures for him,
for any admissible controls of the pursuer, to avoid on a small-time interval
0 ≤ t ≤ θ0 ( θ0 > 0 is a constant) the collision M with the corresponding
trajectory z (t) starting at point z0 = z(0)∈ Rn \M. Then, based on the local
deviation, the evader an algorithm for avoiding the trajectory z (t) from M
for all t ≥ 0 is constructed (global avoiding). At the stage of local avoidance,
the main task is to prove that the distance ρ (z (t) , M ) from the moving point
z (t) of the trajectory to the terminal set is positive for all t ≥ 0, then for
 4.2 Investigation of Nonlinear Hölder Type Integral Inequality 105

these t z (t) ∈/ M , and thus, global avoidance is possible. However, it can be

tough to prove inequality ρ (z (t) , M ) > 0, especially for nonlinear systems.
Therefore, for the distances, they try to obtain lower bounds. Namely, at this
stage, it becomes necessary to use integral inequalities. Let us explain this. If
the players have chosen controls u (t) ∈ P and vc (t) ∈ Q, t ∈ [0, 1], then to
these controls and a point z0 ∈ / M corresponds the solution of system (4.16)
with the initial condition z (0)= z0 , which has the form [8,9]:
t
z (t) = z0 + f (z (s) , u (s) , v (s)) ds.
0

Further, based on his advantage and Hölder condition (4.17), the evading
player constructs his control in such a way that provides him the inequality
t
α
|z (t) − z0 | ≤ d1 (1 + | z 0 | )t + ( d2 +Lα |z (s) − z 0 |α ) ds, (4.18)
0

where constants d1 and d2 are positive, 0 < α ≤ 1.

Inequality (4.18) is a nonlinear integral inequality of the Hölder type
(4.7). With its help, as well as other conditions of the avoidance prob-
lem (4.16–4.17), the main inequality in the theory of differential games of
avoiding encounter is obtained [32,39], which in the general case has the
form
ρ (z (t) , M ) ≥ γtk , (4.19)
where t ∈ [0, θ0 ], γ, k and θ0 are positive constants are effectively calculated
from the initial data of the game (4.16).
Inequality (4.19) ensures that the trajectory does not hit the terminal set
on some “small” time interval. Then, the possibility of global avoidance
(evasion) from the point z0 ∈ / M is proved for all t ≥ 0. Note that even
in the cases of applying the linear integral Grönwall–Bellman inequality [1,
2, 10], obtaining lower bounds of type (4.19) involves extensive calculations
and proofs [32, 39].

4.2.3 Comparison of Results

1. If in (4.7) we assume n = 1 and 1) α1 = 0 or α1 = 1; b)
c1 (t) ≡ 0 for t ≥ t0 , then each case leads to the consideration of well-
known linear integral inequalities and estimates of Grönwall–Bellman
type [1, 2].
106 Nonlinear Integral Inequalities and Differential Games

2. The case when in (4.7) n= 1, a (t) ≡0, f (t) ≡f0 ≥0 (f0 is a constant
for all t≥t0 ), α1 ≥ 0, is studied in [7] (if α1 =0 or α1 =1 we have
nonlinear case). It is necessary to note, that the proof in [7] does not
extend in the general case for the proof of the inequality (4.7).
3. The case when in (4.7) n = 1, a (t) ≡ 0, b (t) ≡ 0, 0 ≤ t ≤ 1,
0 < α1 < 1, considered in [7,12,13].

4.2.4 Remarks

Remark 4.1. First, Hölder type inequality was considered in [6] and investi-
gated in [12,13,58].

Remark 4.2. In many papers, as an integrand in (4.7), a more general function

is considered that has the property of monotonicity on u(·) of the function
being estimated and some other additional conditions. As a result, we obtain
general estimates in terms of inverse functions or mappings [10,11], which
are not constructive and are not adapted to the study of avoidance problems
in the theory of differential games.

Remark 4.3. Inequality (4.8) and the constant h0 are especially important
for obtaining a lower bound for the distance from a moving point of the
trajectory (4.17) of a conflict-controlled dynamical system to a given terminal
set [32–35,37,39].

Remark 4.4. The problem of approximate calculation of the root ξ0 of the

function z (ξ) is solved in [14, 15] for n= 1 only.

4.3 Conclusion
The main aim of this work is to investigate new nonlinear Hölder type integral
inequalities, generalized well-known integral inequalities [6, 12, 13].
New integral inequalities of the Hölder type are proved, and corre-
sponding solutions of nonlinear integral inequalities of the Hölder type are
obtained, which generalized the well-known [6, 12 ,13]. The results on
integral inequalities presented in this chapter allow for solving various control
problems [14–18] and are applicable to solving new nonlinear problems in the
theory of differential avoiding games. They may also be used for constructing
new control functions(strategies) of players in differential games of avoiding
encounters [58, 59]. Further, the received results make it possible to apply
 References 107

them to research solutions of nonlinear differential equations (existence,

uniqueness, continuity, continuous dependence, etc.).
A short excursion into the historical domain of the development of
integral inequalities is represented.

Acknowledgments
The author is grateful to the Academician of the National Academy of Sci-
ences of Ukraine, Professor A.A. Chikrii, for his kind encouragement of this
work. He also thanks to the Organizing Committee of AUTOMATION2020
for the invitation to participate in this high-level event.

References
[1] T.N. Grönwall, ‘Note on the derivatives with respect to a parameter of
the solutions of a system of differential equations,’ Ann. of Math. 20: 2
(1919), pp. 292–296. DOI 10.2307/1967124.
[2] R. Bellman. ‘The stability of solutions of linear differential equations,’
Duke Math. J. 10(1943), pp. 643–647.
[3] P. Hartman, Ordinary Differential Equations, John Wiley & Sons,
NewYork- London-Sydney, 1964, 720 p.
[4] G. Peano. Sull’ integrabilitá delle equazione differenzialli di primo
ordine, Atti. R.Accad.Torino, 21(1885/18886), 667–685 [III.2.4].
[5] I. Bihari, ‘A generalization of a lemma of Bellman and its application to
uniqueness problems of differential equations,’ Acta Math. Acad. Sci.
Hungr. 7:1(1956), pp. 71–94.
[6] A.I. Perov, ‘On Integral Inequalities,’ Proc. of the Seminar on Functional
Analysis, Voronej St. Univ. − Issue 5. − 1957.
[7] S.A. Chaplygin, ‘A new method of approximate integration of differen-
tial equations,’ Trudy TsAGI 130 (1932), pp. 5–17.
[8] E. Coddington and N. Levinson, ‘Theory of Ordinary Differential
Equations’, McGraw-Hill Book Company, Inc. New York–Toronto–
London,1955, − 474 p. (Russian Transl., Moscow: IL.1958).
[9] L.S. Pontrjagin, ‘Ordinary differential equations,’ −M.: Nauka, 1970, −
332 p.
[10] D. Bainov, P. Simeonov, ‘ Integral Inequalities and Applications.’ Dor-
drecht, Springer, Mathematics and its Applications (East European
Series, vol. 57), 1992, −256p.
108 Nonlinear Integral Inequalities and Differential Games

[11] A. N. Filatov, L.V. Sharova, ‘Integral inequalities and the theory of

nonlinear oscillations,’ Nauka, Moscow, 1976.
[12] Sh. Gamidov, ‘Some integral inequalities for boundary value prob-
lems for differential equations,’ Differential Equations, 5: 3 (1969),
pp. 463–472.
[13] A.A. Martynyuk, R.Gutovski, ‘Integral inequalities and stability of
motion,’ Kyiv: Nauk?va Dumka, −1979.
[14] V.M. Kuntsevich et al. (Eds), Control Systems: Theory and Applica-
tions.Series in Automation, Control and Robotics, River Publishers,
Gistrup, Delft, 2018.
[15] Y.P. Kondratenko, V.M. Kuntsevich, A.A. Chikrii, V.F. Gubarev
(Eds). Advanced Control Systems: Theory and Applications. Series in
Automation, Control and Robotics, River Publishers, Gistrup, 2021.
[16] Y.P. Kondratenko, A.A. Chikrii, V.F. Gubarev, J. Kacprzyk
(Eds). Advanced Control Techniques in Complex Engineering Systems:
Theory and Applications. Dedicated to Professor Vsevolod M.
Kuntsevich. Studies in Systems, Decision and Control, Vol. 203. Cham:
Springer Nature Switzerland AG, 2019.
[17] V.M. Kuntsevich, ‘Estimation of Impact of Bounded Perturbations on
Nonlinear Descrete Systems,’ In: Kuntsevich, V.M. et al. (Eds). Control
Systems: Theory and Applications. Series in Automation, Control and
Robotics, River Publishers, Gistrup, Delft, 2018, pp. 3–15.
[18] J. Kacprzyk, et al. ‘A Status Quo Biased Multistage Decision Model for
Regional Agricultural Socioeconomic Planning Under Fuzzy Informa-
tion,’ In:Kondratenko, Y.P., Chikrii, A.A., Gubarev, V.F., Kacprzyk, J.
(Eds) Advanced Control Techniques in Complex Engineering Systems:
Theory and Applications. Dedicated to Prof. V.M.Kuntsevich. Studies
in Systems, Decision and Control, Vol. 203. Cham: Springer Nature
Switzerland AG, 2019, pp. 201–226. DOI: https://doi.org/10.1007/
978-3-030-21927-7_10
[19] R.Isaacs, ‘Differential Games’, John Wiley and Sons, Inc., New York–
London–Sydney, 1965, – 480 p.
[20] L.S. Pontrjagin. ‘Selected Proceedings’− M.: MAKS Press, 2004, −552
p.
[21] N.N. Krasovskiĭ, A.I. Subbotin, ‘Closed-loop differential games,’
Springer–Verlag, Berlin and New-York, 1980, −456 p.
[22] A.B. Kurzhanskiĭ, ‘Control and Observation under Uncertainty,’ −M.:
Nauka, 1977, −392 p.
 References 109

[23] L.A. Petrosjan, ‘On a set of differential games of survive in the space
Rn ’ Dokl. Acad.Nauk SSSR, 1965, V. 161, No. 1, –P.52-54.
[24] M.I. Zelikin, N.T. Tynjanskiĭ, ‘Determinate differential games’,
Usp.Mat. Nauk, 1965, V. 20. No. 4. −P.151–157.
[25] Yu.S. Osipov, ‘Alternative in differential-difference game,’ Dokl. Soviet
Acad.Sc., 1971, V. 197, No. 5, − P. 1022–1025.
[26] A.I. Subbotin, A.G. Chentsov, ‘Optimization of Guarantee in Control
Problems’, -M.: Nauka, 1981, − 288p.
[27] B.N. Pshenichnyĭ, V.V. Ostapenko, ‘Differential games’, Kiiv: Naukova
Dumka, 1992, −261p.
[28] A.A. C̆hikriĭ, ‘Conflict controlled processes’, Boston-London-
Dordrecht, Kluwer Acad.Publ., 1997, − 424 p.
[29] M. Sh, Mamatov, Kh. Kh. Sobirov, ‘To the theory of differential games
of pursuit on position,’ Results of Science and Technics, Modern math.
and its Appl.Subject Review, 2018, v.144, −P. 39–46.
[30] M. Sh. Mamatov, M. Tukhtasinov, ‘On a Problem of Pursuit in Dis-
tributed Control Systems,’ Cybernetics and System Analysis, Kyiv,
2009, No. 2, − P. 153–158.
[31] L.S. Pontrjagin, E.F. Mi˘s c̆enko, ‘A problem of evasion of one controlled
object from an other,’ 1969, Dokl. Acad. Nauk SSSR, V. 189, No. 4,
–P. 721–723.
[32] L.S. Pontrjagin, ‘A linear differential evasion game’, Trudy
Mat.Inst.Steklov, 19 71, V. 112, −P. 30–63.
[33] E.F. Mi˘s c̆enko, N, Satimov, ‘The evasion problem in differential games
with nonlinear controls,’ Differencialnie Uravneniya, V. 9, 10, 1973.
−P.1792-1797.
[34] E.F. Mi˘s c̆enko, M.S. Nikolskiĭ, N. Satimov, ‘The problem of avoiding
encounter in N-person differential games’, Proc. of the Steklov Institute
of Mathematics, 1980, Iss. 1, −P. 111–136.
[35] N.Satimov, ‘On a one way of avoiding encounter in differential games,’
Matematicheskiy sbornik, 99(141), 1976, −P. 280–293.
[36] F.L. Chernousko, A.A. Melikyan, ‘Game Problems of Control and
Search’, – Nauka, 1978,−270 p.
[37] P.B. Gusyatnikov, L.P.Yugay, ‘On an evasion problem in nonlinear
differential games with terminal set of compound structure,’ Izvestiya
Acad. Nauk SSSR, 1977, Techn.kibernetika, No. 2, −P. 8–13.
[38] R.V. Gamkrelidze, G.L. Kharatshvili, ‘A differential game of evasion
with nonlinear control’ SIAM J. Control, May 1974, Vol. 12, No 2,
–P. 332–349.
110 Nonlinear Integral Inequalities and Differential Games

[39] N.Yu. Satimov, B.B. Rikhsiev, ‘Methods of solving of avoiding

encounter problem in mathematical theory of control,’ Fan, Tashkent,
2000, −176 p.
[40] N.Yu. Satimov, ‘To the theory of differential evasion games’ Mat.
Sbornik,1977, V. 103, No. 3. – P. 430–444.
[41] N. Satimov, ‘Composition of differential equation on its solutions and
proof of one lemma of L.S. Pontrjagin,’ Diff. Equations, 1978, V. 14,
No. 7,− P. 1208–1214.
[42] J. Yong, ‘Evasion without superiority,’ J. of Math. Anal. and Appl.,
1988, V. 134, No. 1, −P. 116–124.
[43] V.N. Ushakov, ‘Extremal strategies in differential games with integral
constraints,’ J. Appl. Math.Mech., 1972, Vol. 36, No 1, −P. 12–19. DOI:
10.1016/0021-8928(72)90076-7
[44] N. Satimov, A. Fazilov, A. Khamdamov, ‘On pursuit and evasion
problems in differential and discrete games of many persons with
integral restrictions,’ Differential equations, 1984, Vol. 20, No. 8,
− P. 1388–1396.
[45] N. Mamadaliev, ‘On a one Pursuit Problem with Integral Restrictions
on Controls of Players’, Siberian Math. Journal, Jan.-Febr., Novosibirsk,
2015,V.56, No. 1, −P. 129–148.
[46] N. Mamadaliev, ‘Linear Differential Games with Integral Restrictions
on Controls and Delay,’ Mathematical Notes, 2012, No. 5, - P. 750–760.
[47] L.D. Bercovitz, W.H.Fleming, ‘On differential games with integral pay-
off,’ Contributions to the theory of games, Ann. of Math. Studies, No 3,
1957, −P. 413–435.
[48] B.T. Samatov, ‘On a Pursuit-Evasion Problem under a Linear Change
of the Pursuer Resource,’ Siberian Advances in Mathematics, Allerton
Press, Inc. Springer, New York, 23(4), 2013, − P. 294–302.
[49] B.T. Samatov, E.T. Umarov, ‘Differential Games with Constraints of
the Gronwall Type,’ Abstracts of the VI International Sc. Confer-
ence “ Modern Problems of the Applied Mathematics and Information
Technology–Al–Khorezmiy 2018”, Tashkent, September, 13-15, 2018,
–P. 183–184.
[50] B.Samatov. G. Ibragimov, I. Khodjibaeva, ‘Pursuit–Evasion Differential
Games with Gronwall-Type Constraints on Controls,’ Ural Mathemati-
cal Journal, Vol. 6, No 2, 2020, -P. 95–107.
[51] A.O. Chikrii, G.T. Chikrii. Matrix resolving functions in dynamic
games of approach.Cybernetics and System Analysis, 2014, 50(2),
pp. 201–217.
 References 111

[52] L.P. Yugay, ‘Linear differential game without superiority,’ Annals of

differential equations, China, 1992, No. 2, −P. 158–163.
[53] L.P. Yugay, ‘On an optimization of evasion parameters in minmax
differential games,’ Bulletin of Korean Mathematical Society, Vol. 34,
No. 4, November 1997, P. 495–508.
[54] L.P. Yugay, ‘Avoiding from discrete terminal set by geometrical restric-
tions on controls,’ J.Problems of Controls and Informatics, Kyiv, Inst.
of Cyb. of NAS of Ukraine named after V.M. Glushkov, 1998, No. 1,
– P. 4–11.
[55] L.P. Yugay, ‘To the Problem of Pendulum’s Swinging,’ Proc. of
13-th Int’l Conference “Stability and Oscillation of Nonlinear Con-
trol Systems” (Pyatnitsky’s Conference), Moscow, 1-3 of June 2016,
−P. 429–432.
[56] L.P. Yugay, ‘L.S. Pontryagin’s lemma in differential evasion games’,
Optimal Control and Differential Games: Proc. of Int’l Conf. dedicated
to the 110th - anniversary of L.S. Pontrjagin, Dec. 12–18, 2018, Moscow,
Russia, P. 295–296.
[57] L.P. Yugay, ‘Differential Evasion Game with Locally Inertial Controls,’
Bulletin of L.N. Gumilyov Eurasian National University (Kazakhstan),
Mathematics, Computer Science, Mechanics Series, no. 4, vol. 129,
2019, −P. 54–66.
[58] L.P. Yugay, On Nonlinear Integral Inequality in Collision Avoid-
ing Problems.Abstracts of the XXVI-th International Conference
AUTOMATION-2020 devoted to the memory of Academicians L.S.
Pontryagin, N.N. Krasovsky and B.N. Pshenichny, Kyiv, Ukraine,
13-15.10.2020, –P.142.
[59] L.P. Yugay, ‘The Problem of Trajectories Avoiding a Sparse Terminal
Set,’ Journal: Doklady of Russian Academy of Sciences, Mathematics,
Springer, Pleiades Publishing, Ltd., 2020, Vol.102, no. 3, − P. 538–541.
DOI: 10.1134/S1064562420060228
 5
Principle of Time Dilation in Game
Problems of Dynamics

G.Ts. Chikrii

Glushkov Institute of Cybernetics of the National Academy of Sciences

of Ukraine, 40 Glushkov Avenue, 03187, Kyiv-187, Ukraine
E-mail: [email protected]

Abstract
In this chapter, we consider the linear differential game of approaching a
cylindrical terminal set. The case when classic Pontryagin’s condition does
not hold is studied. Instead, a considerably weaker condition, incorporating
the function of time dilation, is introduced. This makes it possible to expand
the range of problems that can be solved analytically, by using information
about the enemy’s control in the past.
The study is carried out in the framework of the Method of Resolving
Functions. The gist of the method consists in constructing certain scalar
function associated with the parameters of the conflict-controlled process and
characterizing the gain of the first player at each moment. If the total gain
achieves the predetermined value then this means that the game terminates in
a certain guaranteed time. The pursuer’s control, realizing the game goal, is
constructed based on the Filippov–Castaing theorem on measurable choice.
To illustrate the developed scheme, we analyze in detail the problem of
approaching two controlled oscillation systems in geometric coordinates.

Keywords: Differential game, conflict-controlled process, set-valued map-

ping, modified Pontryagin’s condition, function of time dilation, mathemati-
cal pendulum.

113
114 Principle of Time Dilation in Game Problems of Dynamics

5.1 Introduction
The mathematical theory of conflict-controlled processes (differential games,
dynamic games) has at its disposal a wide range of fundamental methods
to study various nature-controlled processes which function in conditions of
conflict and uncertainty.
Attempts to construct optimal behavior of the opposing sides when solv-
ing game dynamic problems inevitably lead to the ideology of dynamic pro-
gramming, closely related to the Hamilton–Jacobi—Bellman–Isaacs equa-
tion, the main equation in the theory differential games [1]. In a slightly
different form, namely in the terms of sets, these ideas are reflected in the
Pontryagin–Pshenichnyi backward procedures [2, 3]. These are the method
of alternated integral and the method of semi-group operators Tε .
According to the aforementioned ideology, in the course of the pro-
cess of counter-acting at each moment, one of the parties maximizes and
another minimizes the payoff functional. Thus, optimization of the conflict-
controlled process is reduced to calculating continual maxmin or minmax of
the functional, depending on the current information available to each of the
parties.
The desire to find optimal solutions for counter-acting parties in the
game problems of dynamics encounters great difficulties of mathematical
nature. Therefore, a number of effective mathematical methods have been
created for deciding dynamic games, which provide a guaranteed result and
give sufficient conditions for the goal achievement without worrying about
optimality, which is quite justified from the practical point of view.
These are the First Direct Method of L.S. Pontryagin [2], the Rule of
Extremal Aiming of N.N. Krasovskii [4, 5], the Method of Programmed
Iterations [6], and the Method of Resolving Functions [7]. Adjacent to them
are the works [8, 9]. Various applications of game methods to attack applied
problems are contained in the books [10, 11].
In the course of a conflict-controlled process, information about the cur-
rent state of the process, as a rule, delays. One of the reasons is that a certain
time is required to process it. The author of this chapter developed the method
of reducing the approach problem with delay of information to the equivalent
problem with complete information, but slightly changed dynamics [12, 13],
followed by application of the known classic methods.
When performing such important tasks as, for example, interception of a
moving target, in order to achieve the goal, a certain advantage of the pursuer
in terms of control resources and (or) dynamic capabilities is necessary.
 5.1 Introduction 115

Classic Pontryagin’s condition [2] plays this role in general problem state-
ments. There exist various forms of this condition [7]. One of them consists
in incorporating the solid part of the terminal set into Pontryagin’s condition.
Another way concerns the suppression of the evader’s control resource by the
way of multiplying it by a certain matrix function and subsequent subtraction
of the used resource from the terminal set. In so doing, Pontryagin’s condition
falls into two conditions. The latter technique is appropriate, for example, in
the case when the opposing objects have different inertia.
However, in many cases of conflict withstanding Pontryagin’s condition
does not hold, e.g. the problems of soft meeting and the pursuit problems for
oscillatory processes.
M.S. Nikolskij and D. Zonnevend [14, 15] suggested weakening Pon-
tryagin’s condition by the way of incorporating into it a certain function,
later on, called the function of time dilation. The establishment of the close
relation of the modified condition with the passage from the original game
with complete information to an auxiliary game with special kind information
delay [13] gave impetus to the development of an efficient approach the so-
called principle of time dilation for solving problems for which Pontryagin’s
condition does not hold.
Essentially, the transition is made from the original game with complete
information to the game with the same dynamics and the terminal set, yet
with a special kind of information delay, decreasing as the game trajectory
approaches the terminal set and vanishing as it hits the target... An important
point is that Pontryagin’s condition for the latter game involves the time
function of time dilation.
We extend the concept of the function of time dilation to the class of
functions of bounded variation. This allows applying the above-mentioned
technique to solving a wide range of game problems for the processes,
described by the systems of second-order differential equations [16, 17] and
for the problems of soft meeting [18].
In this chapter, we apply the principle of time dilation to the game
constructions based on the method of resolving functions [7, 19] This method
is closely related with the Minkowski’ inverse functionals, constructed for
special set-valued mappings. An attractive feature of the method of resolving
functions is that it provides the full theoretical substantiation of the classic
rule of parallel pursuit and the approach along the ray [20]. Moreover,
when substantiating the game constructions and obtaining meaningful results
on their basis, the modern technique of set-valued mappings is effectively
used.
116 Principle of Time Dilation in Game Problems of Dynamics

The major contributor to the creation of the mentioned method was

a special minmaximin function in the theory of avoidance of a group of
pursuers [21]. It allowed BN Pshenichnyi to formalize the situation of the
environment [22]. Subsequently, the theory of group pursuit was gathering
force [7, 23, 24]. The broad possibilities of the method of resolving functions
are illustrated on the problem of successive pursuit, the commercial traveler
type problem by using Apollonian circles [25], and in the problems with state
constraints [7].
The accumulative principle inherent in the method looks quite natural.
Its versatility allows, in a unified scheme, to encompass conflict-controlled
processes described by functional differential equations [26, 27], partial
differential equations [28], equations with fractional derivatives [29], impulse
systems [30], non-stationary processes [31, 32], and stochastic systems [33].
Therewith, the approach conditions allow to apply of the quasi- and strobo-
scopic strategies [4, 8] in the course of the game, based on the theorems on
measurable choice [34].

5.2 Statement of the Game Problem. Classic and Modified

Pontryagin’s Condition
Let the motion of a conflict-controlled object in the finite-dimensional space
Rn be described by the system of differential equations

ż = Az + u − v, (5.1)

A is a quadratic matrix of order n.

Parameters of the players’ controls u and v are picked up from the
domains of players controls U and V , which are compacts from Rn .
Besides, a terminal set M ∗ , having a cylindrical form, is given:

M ∗ = M0 + M, (5.2)

where M0 is a linear subspace in Rn , M - a compact from L, the orthogonal

complement to M0 in Rn .
The goal of the first player (the pursuer) is in the shortest time to bring
a trajectory of the process (5.1) to the terminal set (5.2) with the help of an
appropriate choice of a control parameter
The second player (the evader) strives, by choosing parameter v, to avoid
meeting the trajectory of the process (5.1) with the set M ∗ within a finite time
or, if it is impossible, to postpone this meeting as much as possible.
5.2 Statement of the Game Problem. Classic and Modified Pontryagin’s Condition 117

To fully formulate the problem of approach, it is necessary to stipulate

what information is available to the players in the process of conflict coun-
teraction. Let’s take the side of the first player and find out what result he can
guarantee himself.
We assume that the second player chooses as admissible controls arbitrary
measurable functions with values from the compact V . Let us denote by
ΩE the set of all admissible controls of the evader.
If at each moment of making a decision t, t ≥ 0, the first player uses
information about the initial state of the process z0 and the prehistory of
control of the evader

vt (·) = {v(s) : v(s) ∈ V, s ∈ [0, t]},

i.e.u(t) = u(z0 , t, vt (·)), then, we say that he applies quasi-strategy.

In the case the pursuer at the current moment t uses only informa-
tion about the initial state z0 and instantaneous control of the evader v(t),
i.e.u(t) = u(z0 , t, v(t)), then they say about counter-control by N.N.
Krasovskii [4], prescribed by stroboscopic strategy by O.Hayek [8].
In this chapter, we discuss using delayed information by the first player t
with the goal of solvability of the original problem of approach. The case of
lack of advantage of the first player in control resources is considered, that is
when Pontryagin’s condition [2] does not hold. Let us proceed to the scheme
of solving the problem (5.1), (5.2) based on the method of resolving functions
[7, 19] and formulation of the above-mentioned condition [2].
Let us denote by π the operator of orthogonal projection acting from
n
R onto L and take into consideration the set-valued mapping W (t) =
πeAt u∗ πeAt v, where ∗ is the operation of geometric subtraction by
− −
Minkowski [2, 7].

Pontryagin’s condition. The set-valued mapping W (t)has non-empty

images for t ∈ [0, +∞).
Under this condition, the scheme of the method of resolving functions
allows obtaining sufficient conditions for the solvability of the game problem
of approach for various classes of strategies [7, 19].
However, there are many problems for which Pontryagin’s condition does
not hold.
In what follows we outline the procedure which makes it feasible to
weaken this condition. It is related to the principle of time dilation [16, 17]
and allows a widening class of conflict-controlled processes, for which the
118 Principle of Time Dilation in Game Problems of Dynamics

problem of approach can be solved in a finite time, in their number, in the

framework of the method of resolving functions.
Modified Pontryagin’s condition. There exists a differentiable monotoni-
cally increasing function I(t),I(t) ≥ t, t ≥ 0, I(0) = 0, such that the
set-valued mapping
˙
WI (t) = πeAt U ∗ I(t)πe AI(t)
V, t ≥ 0,
−

has non-empty images.

5.3 Method Scheme

Since the mapping WI (t) is measurable and closed-valued, then, by the
theorem on measurable choice [34], there exists a measurable selection
γI (t),γI (t) ∈WI (t). Let us fix it for use in the sequel and introduce a function
 I(t)−t  t
At
N (t, u0 (·)) = πe z0 + πe A(I(t)−s)
u0 (s)ds + γI (t − s)ds.
0 0
It depends on the initial state and the control on the initial interval of time
u0 (τ ), u0 (τ ) ∈ U, τ ∈ [0, I(t) − t].
Following the ideology of the method of resolving functions [7, 19], we
analyze the mapping

AI (t, s, v, u0 (·)) = {α ≥ 0 : [WI (t − s, v) − γI (t − s)] α
[M − N (t, u0 (·))] = } (5.3)
where
˙
WI (t, v) = πeAt U − I(t)πe AI(t)
v, t ≥ 0, v ∈ V.
Its support function in the direction +1 is called the resolving function [7]:
αI (t, s, v, u0 (·)) = sup{α : α ∈ AI (t, s, v, u0 (·))}.
From expression (5.3) it follows that, if for some u0 (·) and t, t ≥ 0,
N (t, u0 (·)) ∈ M , then AI (t, s, v, u0 (·)) = [0, +∞) and αI (t, s, v, u0 (·)) =
+∞ for all s ∈ [0, t], v ∈ V.
Let us consider the set
 t
TI (z0 , γI (·)) = {t ≥ 0 : sup inf αI (t, s, v(s), u0 (·))ds ≥ 1},
u0 (·)∈U0t v(·)∈ΩE 0
(5.4)
 5.4 Main Statement 119

where
U0t = {u0 (s) : u0 (s) ∈ U, s ∈ [0, I(t) − t]}.
If the inequality in parentheses (5.4) is not fulfilled for all t ≥ 0, then
we put TI (z0 , γI (·)) = Ø. In what follows, we assume that the exact upper
bound u0 (·) is achieved in (5.4). For example, as a rule, in the case, the set
U contains 0, for simplicity’s sake, control of the first player on the initial
time interval is set identically equal to zero that providing fulfillment of the
inequality in (5.4).
Before proceeding to the formulation and proof of the main result, let
us perform some rearrangements in the representation of the system solution
(5.1), taking into account the availability of information of the first player
about his adversary control choice.
Let t be an arbitrary moment. Using the Cauchy formula, for the chosen
controls of the players, we present o the solution of system (5.1) in projection
onto the subspaceL:
 t  t
πz(t) = πeAt z0 + πeA(t−s) u(s)ds − πeA(t−s) v(s)ds.
0 0

Taking into account the crucial role of the function of time dilation, we
separate the section of initial control of the evader. Then we have
 I(t)−t
πz(I(t)) = πeAI(t) z0 + πeA(I(t)−s) u0 (s)ds+
0
 I(t)  I(t)
+ πeA(I(t)−s) u(s)ds − πeA(I(t)−s) v(s)ds. (5.5)
I(t)−t 0

In order to equalize the limits of integration, we change the variables in the

last two terms of formula (5.5), s by I(t) − t + s and I(t) − I(t − s),
respectively.
As a result, we obtain the following representation of the solution
 I(t)−t
πz(I(t)) = πeAI(t) z0 + πeA(I(t)−s) u0 (s)ds+
0
 t
+ πeA(t−s) u(I(t) − t + s)ds
0
 t
− ˙ − s)ds.
πeAI(t−s) v(I(t) − I(t − s)) · I(t (5.6)
0
120 Principle of Time Dilation in Game Problems of Dynamics

5.4 Main Statement

Theorem Let the modified Pontruagin’s condition with some function I(t) be
satisfied for the conflict-controlled process (5.1), (5.2) and let the set M be
convex.
Then, if for the initial state z0 there exists a measurable selection γI (t) of the
set-valued mapping WI (t), t ≥ 0, such that

TI (z0 , γI (·)) = andT∗ ∈ TI (z0 , γI (·)),

then the trajectory of the process (5.1) can be brought to the terminal set at
the moment I(T∗ ).
Proof. Let us study the process (5.1), (5.2) on the interval [0, I(T∗ )]. On the
initial interval [0, I(T∗ ) − T∗ ], we choose the initial control u0 (·) from the
condition for a maximum of the following expression
 T∗
inf αI (T∗ , s, v(s), u0 (·))ds.
v(·)∈ΩE 0

Then, on the interval (I(T∗ ) − T∗ , I(T∗ )] of the lengthT∗ , we count the

time from 0 to T∗ . Let us consider two cases, namely, N (T∗ , u0 (·)) ∈/ M , and
N (T∗ , u0 (·)) ∈ M , the latter corresponding to the resolving function equal
to +∞.
In the first case, under the scheme of the method of resolving function
[19], we introduce into consideration the controlling function
 t
hI (t) = 1 − αI (T∗ , s, v(I(T∗ ) − I(T∗ − s)), u0 (·)ds,
0
v(s) = v(I(T∗ ) − I(T∗ − s)) ∈ V. (5.7)

Here v(s) is an arbitrary admissible control of the evader.

By virtue of the continuity of a function hI (t), from the definition of the
time T∗ and the theorems of mathematical analysis, it follows that there exists
time t∗ , such that hI (t∗ ) = 0.
We name [0, t∗ ] by the active interval and (t∗ , T∗ ] – by the passive one.
To them correspond the intervals [I(T∗ ) − T∗ , I(T∗ ) − T∗ + t∗ ] and (I(T∗ ) −
T∗ +t∗ , I(T∗ )], respectively. The corresponding representations of solutions at
t = T∗ , not separating the active and passive intervals, are given by formulas
(5.5) and (5.6).
 5.4 Main Statement 121

Let us choose the pursuer control as follows. In the case N (T∗ , u0 (·)) ∈
/
M , we consider the set-valued mapping

U (I(T∗ ) − T∗ + s, v(I(T∗ ) − I(T∗ − s))) = {u(I(T∗ ) − T∗ + s) ∈ U :

πeA(T∗ −s) u(I(T∗ ) − T∗ + s) − πeAI(T∗ −s) v(I(T∗ ) − I(T◦ − s)) · I(T
˙ ∗ − s)
− γI (T∗ − s) ∈∈ ᾱI (T∗ , s, v(I(T∗ ) − I(T∗ − s)), u0 (·))
[M − N (T∗ , u0 (·))]}. (5.8)

Here
ᾱI (T∗ , s, v(I(T∗ ) − I(T∗ − s)), u0 (·)) =

αI (T∗ , s, v(I(T∗ ) − I(T∗ − s)), u0 (·)), s ∈ [0, t∗ ],
=
0, s ∈ (t∗ , T∗ ].
In view of the measurability of this mapping, there exists its selection,
which is a superposition of measurable functions [34]

uI (s) = u(I(T∗ ) − T∗ + s, v(I(T∗ ) − I(T∗ − s)), s ∈ [0, T∗ ]. (5.9)

We assign just this function to be control of the first player. It is easy to

see that the pursuer chooses his control based on delayed information about
the evader’s control since in view of the above-outlined scheme,

I(T∗ ) − T∗ + s ≥ I(T∗ ) − I(T∗ − s).

If, otherwise, N (T∗ , u0 (·)) ∈ M , then control of the first player is chosen
according to a similar scheme, but with the resolving function equal to zero
on the entire interval [0, T∗ ]. This case [19] is said to correspond to the first
direct method of L.S. Pontryagin [2].
Setting t = T∗ and separating the active and the passive segments, we
finally derive the representation of the solution from formula (5.6):
 I(T∗ )−T∗
AI(T∗ )
πz(I(T∗ )) = πe z0 + πeA(I(T∗ )−s) u0 (s)ds+
0
 t∗  T∗
A(T∗ −s)
+ πe u(I(T∗ )−T∗ +s)ds+ πeA(T∗ −s) u(I(T∗ )−T∗ +s)ds−
0 t∗
 T∗
− πeAI(T∗ −s) v(I(T∗ ) − I(T∗ − s) · I(T
˙ ∗ − s)ds.
0
122 Principle of Time Dilation in Game Problems of Dynamics

In the case N (T∗ , u0 (·))∈

¯ M , using the above-described rules of control
selection on the active and passive parts, we obtain

πz(I(T∗ )) ∈ N (T∗ , u0 (·))+

 T∗
+ ᾱI (T∗ , s, v(I(T∗ ) − I(T∗ − s)), u0 (·))[M − N (T∗ , u0 (·))]ds =
0
 T∗
= N (T∗ , u0 (·))[1 − ᾱI (T∗ , s, v(I(T∗ ) − I(T∗ − s)), u0 (·))ds]+
0
 T∗
+ ᾱI (T∗ , s, v(I(T∗ ) − I(T∗ − s)), u0 (·))M ds = M. (5.10)
0

Note that in the last equation of (5.10), the convexity of the solid part of the
terminal set was taken into account.
If, otherwise, N (T∗ , u0 (·)) ∈ M , then, in view of the choice of control
(5.8), (5.9) with the zero resolving function, from (5.10) readily follows the
inclusion
πz(I(T∗ )) ∈ M.

Remark 5.1. The scheme of the method and main assertion are valid for
much more general classes of conflict-controlled processes than described by
(5.1), in particular, for generalized functional-differential systems, including
fractional, impulse, and time-delay systems.
Remark 5.2. The function of time dilation can be discontinuous, but abso-
lutely continuous on the intervals of its continuity [17].

5.5 Example
To illustrate the proposed approach, let us investigate the game problem of
convergence in geometric coordinates of two controlled systems of second-
order, describing the dynamics of the mathematical pendulum. Let their
dynamics be given by the equations

ẍ + a2 x = ρu, x ∈ Rn ,
ÿ + b2 y = σv, y ∈ Rn . (5.11)

Here x and y are geometric coordinates of the players, the control param-
eters are u v, u ≤ 1, v ≤ 1, a2 and b2 are the stiffness control factors, ρ
 5.5 Example 123

and σ – force coefficients, a, b, ρ, σ – positive numbers, a > b, therewith a, b

are proper frequencies of circle oscillations of corresponding systems [35].
The initial positions and velocities of the players are
x(0) = x0 , ẋ(0) = ẋ0 , y(0) = y0 , ẏ(0) = ẏ0 .
With the help of change of variables
z1 = x, z2 = ẋ, z3 = y, z4 = ẏ,
we pass from the system of second-order (5.11) to the system of the first order
ż1 = z2 ,
ż2 = −a2 z 1 + ρu, (5.12)
ż3 = z4 ,
ż4 = −b2 z3 + σv,
z0 = (z10 , z20 , z30 , z40 ), z10 = x0 , z20 = ẋ0 , z30 = y0 , z40 = ẏ0 .
The terminal set takes the form
M0 = {z = (z1 , z2 , z3 , z4 ) : z1 = z3 }, M = {0}.
The matrix of orthogonal projecting is
π = ( E 0 −E 0 )T ,
where 0 and E are unit and zero matrices of ordern, respectively.
The control domains have forms
⎛ ⎞ ⎞⎛
0 0
⎜ ρS ⎟ ⎜ ⎟
U =⎜ ⎟,V = ⎜ 0 ⎟,
⎝ 0 ⎠ ⎝ 0 ⎠
0 σS
where S− the unit ball is centered at the origin.
Meeting of the controlled objects at some moment t, t > 0, testifies
that πz(t) = 0.
The fundamental matrix of (5.12) has the form [35]
⎛ ⎞
cos at · E a sin at · E
1
0 0
⎜ −a sin at · E cos at · E 0 0 ⎟
eAt = ⎜
⎝
⎟,
⎠
0 0 cos bt · E 1
b sin bt · E
0 0 −b sin bt · E cos bt · E
124 Principle of Time Dilation in Game Problems of Dynamics

where ⎛ ⎞
0 E 0 0
⎜ −a2 E 0 0 0 ⎟
A=⎜
⎝ 0
⎟.
0 0 E ⎠
2
0 0 −b E 0
Let us write Pontryagin’s condition [2]. In the case in study, it takes the form
 
[ρ/a |sin at| u − σ/b |sin bt| v] = ρ/a |sin at| S ∗ σ/b |sin bt| S =
v≤1u≤1

= (ρ/a |sin at| − σ/b |sin bt|)S = Ø ∀t ≥ 0.

One can see that for the fulfillment of Pontryagin’s condition it is necessary
that the following inequality be satisfied
ρ/ |sin at| − σ/ |sin bt| ≥ 0∀t ≥ 0.
a b
But it can hold only under very burdensome conditions [7].
Let us write the modified Pontryagin’s condition that incorporates the
function of time dilation I(t)
WI (t) = (ρ/a |sin at| − σ/bI(t)
˙ |sin bI(t)|)S = Ø ∀t ≥ 0.
This condition is true if the following inequality has a place
ρ/ |sin at| − σ/ I(t)
˙ |sin bI(t)| ≥ 0∀t ≥ 0. (5.13)
a b
Let us find the function of time dilation, which satisfies all requirements
mentioned in its definition and provides fulfillment of the inequality (5.13).
We set
a
I(t) = t∀t ≥ 0. (5.14)
b
˙ = a.
Since a > b, then I(t) > t, I(0) > 0, I(t) b
Under additional condition
ρ3 2 ≥ σ3 2 , (5.15)
a b
inequality (5.13) is readily satisfied.
For the function of time dilation (5.14) the set-valued mapping WI (t) has
the form: 3 3
WI (t) = a(ρ a2 − σ b2 ) |sin at| · S. (5.16)
It is easy to see that, under condition (5.15), it has non-empty images.
 5.5 Example 125

Let us proceed to implement the scheme of the method of resolving

functions to the problem at hand.
The set-valued mapping WI (t) contains zero for all t ≥ 0. Therefore, for
the simplicity sake of mathematics, we choose for a measurable selection
γI (t) of the mapping WI (t), the function identically equal to zero, i.e.
γI (t) ≡ 0. Moreover, with the same goal for simplicity, since the control
domain of the first player is a unit ball centered at the origin, we set his
starting control u0 (τ ) equal to zero: u0 (τ ) ≡ 0, τ ∈ [0, I(t) − t] for all t > 0.
It should be emphasized that the aforementioned point in time t > 0 has
not yet been found and will be discussed in the sequel. Function N (t, u0 (·))()
takes the following form:

N (t, 0) = πeAt z0 = cos bI(t)y0 − cos aI(t)x0

+ 1 / b sin bI(t)ẏ0 − 1 / a sin bI(t)ẋ0 .

Using formulas (5.16) and (5.3), we obtain

3
WI (t, v) = ρ/a |sin at| S − σa b2 |sin at| v,

AI (t, s, v, 0) = {α ≥ 0 : −αN (t, 0) ∈ WI (t − s, v)}.

In view of the spherical form of the set-valued mapping WI (t, v), the
resolving function,

αI (t, s, v, 0) = sup{α : α ∈ A(t, s, v, 0)},

is the greatest root of the following quadratic equation for α:

 σa 
 
 2 |sin a(t − s)| v − αN (t, 0) = ρ/a |sin a(t − s)| .
b
Because the norm is a square root of the scalar product of vectors, we
square both parts of the above equation and obtain the quadratic equation for
α. Its greatest root stands for the resolving function and has the form
σa
b2
c(t, s)(v, N (t, 0))
αI (t, s, v, 0) = 2 +
N (t, 0)
2 2 1
[{σa
b2
c(t, s)}2 (v, N (t, 0))2 +N (t, 0)2 (ρ2/a2 c2 (t, s)− σb4a c2 (t, s) v2 )] /2
+ ,
N (t, 0)2
c(t, s) = |sin a(t − s)| .
126 Principle of Time Dilation in Game Problems of Dynamics

Minimum in v, v ≤ 1, of this function, is achieved on the element v =

N (t,0)
− N (t,0) and is equal to

(ρ/a − σa
b2
) |sin a(t − s)|
min αI (t, s, v, 0) = .
v≤1 N (t, 0)
This can be verified using the Lagrange multiplier method [36].
Thus, by virtue of (5.4), the issue concerning the existence of a finite time
for the game termination is reduced to the existence of a positive root of the
following equation for t:
 t 4
σ
a (ρ a2 − 2 ) |sin a(t − s)| ds = N (t, 0) . (5.17)
0 b
Here

N (t, 0) = cos bI(t)y0 − cos aI(t)x0 + 1/b sin bI(t)ẏ0 − 1/a sin bI(t)ẋ0 ,

where the function I(t) is given by the formula (5.14).

Since,N (0, 0) = y0 − x0 = 0, then N (t, 0) > 0 for t = 0 and the left
side of (5.17) turns into zero at t = 0. Herewith, N (t, 0) is bounded at t > 0
while the
ρ 3 part of (5.17) is growing without bound as t → +∞.
3 left-hand
σ
Since, a > b2 , then a finite root of equation (5.17) always exists.
2
Let T∗ be a positive root of equation (5.17) and let controls of the first
player on the active and passive parts, separated by the zero of the controlling
function (5.7), are chosen in compliance with the theorem-proof, that is, from
the relationships (5.8), (5.9). Then at the moment T∗ , the conflict-controlled
process (5.12) hits the terminal set for any admissible behavior of the second
player.
Remark 5.3. Similar control problems are analyzed in [37, 38].

5.6 Conclusion
We present a method for solving the game problems of bringing a trajectory
to the terminal set in the case when the classical Pontryagin’s condition is not
satisfied. This became feasible due to the introduction of the function of time
dilation and a weaker modified condition based on it
The method of resolving functions is used as a basis for investigation.
This method allows obtaining sufficient conditions for the approach in a finite
guaranteed time. In so doing, the first player constructs its control based on
 References 127

delayed information on control of the second player, and the choice of control
itself is carried out based on the theorem on measurable choice.
The results are illustrated by the example of the game approach in geo-
metric coordinates of two second-order oscillatory systems, describing the
dynamics of mathematical pendulums.

References
[1] R.F. Isaacs, ‘Differential Games’, New York–London–Sydney: Wiley
Interscience, 479 p., 1965.
[2] L.S. Pontryagin, ‘Selected scientific papers’, 2, Moscow: Nauka, 576
p., 1988. (in Russian).
[3] B.N. Pschenitchnyi, ‘ε-strategies in Differential Games’, Topics in Dif-
ferential Games, New York, London, Amsterdam: North-Holland Publ.
Co., pp. 45–99, 1973.
[4] N.N. Krasovskii, ‘Game Problems on the Encounter of Motions,
Moscow: Nauka, 420 p., 1970 (in Russian).
[5] B.N. Pschenitchnyi, ‘Linear Differential Games’, Avtom, Telemekh.,
p. 65–78, No 1, 1968 (in Russian).
[6] A.I. Subbotin, A.G. Chentsov, ‘Optimization of Guaranteed Result in
Control Problems’, Moscow: Nauka, 286 p., 1981 (in Russian).
[7] A.A. Chikrii, ‘Conflict controlled processes’, Boston; London, Dor-
drecht: Springer Science and Business Media, 424 p., 2013.
[8] O. Hajek, ‘Pursuit games’, New York: Academic Press, 266 p., 1975.
[9] M.S. Nikolskii, ‘L.S. Pontryagin’s First Direct Method in Differential
Games’, Izdat. Gos.Univ., Moscow, 65 p., 1984 (in Russian).
[10] G. Siouris, ‘Missile Guidance and Control Systems’, New York:
Springer-Verlag, 666 p., 2004.
[11] G. Siouris, ‘Aerospace Avionics Systems: A Modern Synthesis’, San
Diego: Academic Press, 466 p., 1993.
[12] G.Ts. Chikrii, ‘An approach to solution of linear differential games with
variable information delay’, Journal of Automation and Information
Sciences: Begell House Inc., p. 163–170, No 3, 4, vol. 27, 1995.
[13] G.Ts. Chikrii, ‘Using the effect of information delay in differential
pursuit games’, Cybernetics and Systems Analysis, p. 233–245, No. 2,
vol. 43, 2007.
[14] M.S. Nikolskij, ‘Application of the first direct method in the linear
differential games’, Izvestia Acad. Nauk SSSR, p. 51–56, vol. 10, 1972
(in Russian).
128 Principle of Time Dilation in Game Problems of Dynamics

[15] D. Zonnevend, ‘On one method of pursuit’, Doklady Akademii Nauk

SSSR, p. 1296–1299, vol. 204, No 6, 1972 (in Russian).
[16] G.Ts. Chikrii, ‘Principle of time stretching in evolutionary games of
approach’, Journal of Automation and Information Science, Begell
House, Inc., p. 12–26, vol. 48, No. 5, 2016.
[17] G.Ts. Chikrii, ‘Principle of time stretching for Motion Control in Condi-
tion of Conflict’, Advanced Control Systems: Theory and Applications,
River Publishers Series in Automation, Control and Robotics, p. 53–82,
2021.
[18] G.Ts. Chikrii, ‘One Approach to solution of complex game problems
for some quasilinear evolutionary systems’, International Journal of
Mathematics, Game Theory and Algebra, p. 307–314, vol. 14, 2004.
[19] A.A. Chikrii, ‘An analytical method in dynamic pursuit games’, Pro-
ceedings of the Steklov Institute of Mathematics, p. 69–85, vol. 271,
2010.
[20] Locke S. Arthur, Guidance, D.Van Nostrand Company, Inc. Princeton,
776 p., 1955.
[21] A.A. Chikrii, ‘Linear problem of avoiding several pursuers’, Engineer-
ing Cybernetics, p. 38–42, vol. 14, No 4, 1976.
[22] B.N. Pschenitchnyi, ‘Simple Pursuit by Several Objects’, Kibernetika,
p. 145–146, No 3, 1976 (in Russian).
[23] B.N. Pschenitchnyi, A.A. Chikrii, J.S. Rappoport, ‘Group Pursuit in
Differential Games’, Opt. Invest. Stat., Germany, p. 13–27, 1982 (in
Russian).
[24] A.A. Chikrii, ‘Differential Games with Many Pursuers’, Mathematical
Control Theory, Banach Center Publ., PWN, Warsaw, p. 81–107, vol.
14, 1985 (in Russian).
[25] A.A. Chikrii, S.F. Kalashnikova, ‘Pursuit of a group of evaders by a
single controlled object’, Cybernetics, pp. 437–445, No 4, 1987.
[26] A.A. Chikrii, J.S. Rappoport, K.A. Chikrii, ‘Multivalued mappings and
their selectors in the theory of conflict-controlled processes’, Cybernet-
ics and Systems Analysis, p. 719–730, vol. 43, No 5, 2007.
[27] A.A. Chikrii, V.K. Chikrii, ‘Image Structure of Multivalued Mappings
in Game Problems of Motion Control’, Journal of Automation and
Information Sciences, pp. 20–35 vol. 48, No. 3, 2016.
[28] L.A. Vlasenko, A.G. Rutkas, A.A. Chikrii, ‘On a differential game in
an abstract parabolic system’, Proceedings of the Steklov Institute of
Mathematics, pp. 254–269, vol. 293, Issue 1 Supplement, 2016.
 References 129

[29] A.A. Chikrii, S.D. Eidelman, ‘Game problems for fractional quasilinear
systems’, Computers and Mathematics with Applications, p. 835–851,
vol. 44, No. 7, 2002.
[30] A.G. Nakonechnyi, E.A. Kapustyan, A.A. Chikrii, ‘Control of impulse
system in conflict situation’, Journal of Automation and Information
Science, p. 54–63, No. 9, 2019.
[31] L.V. Baranovskaya, A.A. Chikrii, Al.A. Chikrii, ‘Inverse Minkowski
functional in a nonstationary problem of a group pursuit’, Journal of
Computer and Systems Sciences International, p. 101–106, vol. 36,
No. 1, 1997.
[32] L.V. Baranovskaya, Al.A. Chikrii, ‘Game problems for a class of
hereditary systems’, Journal of Automation and Information Science,
p. 87–97, vol. 29, No 2, 1997.
[33] L.A. Vlasenko, A.G. Rutkas, A.A. Chikrii, ‘On a differential game in a
stochastic system’, Proceedings of the Steklov Institute of Mathematics,
p. 185–198, vol. 309, suppl.1, 2020.
[34] J.-P. Aubin, H. Frankowska, ‘Set-valued analysis’, Boston; Basel;
Berlin: Birkhauser, 1990.
[35] N.V. Vasilenko, ‘Theory of Oscillations, Kiev: Vyshcha Shkola, 430
p., 1992 (in Russian).
[36] A.D. Joffe, V.M. Tikhomirov, ‘Theory of Extremal Problems’, North-
Holland, Amsterdam, 480 p., 1979.
[37] V.M. Kuntsevich et al (Eds). Control Systems: Theory and Applications.
Series in Automation, Control and Robotics, River Publishers, Gistrrup,
Delft, 2018.
[38] Y.P. Kondratenko, V.M. Kuntsevich, A.A. Chikrii, V.F. Gubarev
(Eds). Advanced Control Systems: Theory and Applications, Series in
Automation, Control and Robotics, River Publishers, Gistrrup, 2021.
 6
Method of Upper and Lower Resolving
Functions for Pursuit Differential-difference
Games with Pure Delay

Lesia Baranovska

Institute of Applied System Analysis of the National Technical University

of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”,
37 Peremohy av., Kyiv, 03056, Ukraine
E-mail: [email protected]

Abstract
A modification of the method of resolving functions for differential-
difference games with pure delay is considered. The lower and upper resolv-
ing functions are introduced. For the local game of approach, the scheme of
a method is constructed and sufficient conditions for the completion of the
game are formulated.

Keywords: Differential-difference games, differential games, dynamic

games, games of approach, conflict-controlled processes, game theory.

6.1 Introduction
Differential games is a section of mathematical control theory [1–3] studying
the manipulation of moving objects operated under conditions of uncertainty
and conflict. We consider the differential-difference games of approach. In the
theory of differential games, along with Isaacs ideology [4], several methods
have been developed that provide a guaranteed result. The most famous
methods are the first direct method of L.S. Pontryagin [5, 6], the method
of extreme aiming of M.M. Krasovskii [7, 8], and the method of resolving
functions of A.O. Chikrii [9–15]. In this chapter, the method of resolving

131
132 Method of Upper and Lower Resolving Functions

functions will be used. The first direct method of L.S. Pontryagin and the
method of resolving functions give a theoretical justification for the classical
rule of parallel pursuit. Most processes depend not only on the current state,
but they have essential knowledge of the prehistory of the state of the process,
which can’t be ignored. Today, systems of differential-difference equations
are increasingly used in control theory. The method scheme for differential-
difference games is developed in works [16–19]. In [20–22], the problem
of approach for a group of pursuers and a single evader is considered.
The scheme of the method of resolving functions for differential-difference
systems of neutral type was developed in [23]. For objects with different
inertia in [24], a modification of the Pontryagin condition by introducing the
body part of the terminal set is proposed. Differential games in case of failure
of control devices are considered in the works [25–27].
In this chapter, we first introduce the upper and lower resolving functions
of different types [28] in the method of resolving functions for the differential
game of approach with pure delay, which allows developing the method even
if the Pontryagin condition is not fulfilled. Sufficient conditions have been
obtained to complete the game.

6.2 Statement of the Problem

n
Denote by 2R a set of all subsets of space Rn , by K (Rn ) a set of all
nonempty compacts in Rn , and by K (Rn ) a set of all nonempty convex
compacts in Rn . By a set-valued mapping is meant a mapping acting from
n
Rn to 2R [9] and transforming each element x ∈ Rn into a set in Rn .
Consider a controlled object whose dynamics is described by the linear
differential-difference system with pure delay in a Euclidean space Rn
ż(t) = Bz(t − τ ) + ϕ(u, v), z ∈ Rn , u ∈ U, v ∈ V, (6.1)
where B is a square constant matrix of order n; z is the state vector, involving
geometric coordinates, velocities, accelerations of the pursuer, and evader;
U, V ∈ K (Rn ) ;ϕ : U × V → Rn is jointly continuous in its variables.
The initial condition
z(t) = z 0 (t) , −τ ≤ t ≤ 0, (6.2)
is absolutely continuous on [−τ, 0].
The controls u(s) and v(s) that are allowed are the Lebesgue measurable
functions.
 6.3 Scheme of the Method 133

Denote
ΩU = {u(s) : u(s) ∈ U, s ∈ [0, +∞)} ,
ΩV = {v(s) : v(s) ∈ V, s ∈ [0, +∞)} .
The function u(·) ∈ ΩU (v(·) ∈ ΩV ) ,will be called an open-loop pursuer’s
(evader’s) control. This control is chosen by the pursuer (evader) based on
knowledge of the initial condition.
The function vt (·) = {v(s) : s ∈ [0, t], v(·) ∈ ΩV } , will be called
prehistory of the control of evader at a time t, t ≥ 0 [9].
) We define * the pursuer’s “quasistrategy” [29] as a mapping
U t, z 0 (·), vt (·) . It assigns a Lebesgue measurable function to each
moment t ≥ 0, initial condition (6.2) and arbitrary prehistory vt (·) of the
control of evader, taking its values in control domain U:
) *
u(t) = U t, z 0 (·), vt (·) , t ≥ 0, (6.3)

Strobostrophic strategies [30] are a special case of ‘quasistrategies’. The

counter control of the pursuer is based on information on its initial state (6.2)
of the process (6.1) and the instantaneous value of the control of the evader
) *
u(t) = U z 0 (·), v(t) , t ≥ 0. (6.4)

The terminal set has a cylindrical form [9]:

M ∗ = M0 + M, (6.5)

where M0 is a linear subspace inRn and M is a compact set from the

orthogonal complement (L) of M0 in Rn .
The goal of the pursuer (u) is in the shortest time to bring a trajectory of
the process (6.1), (6.2) to a terminal set (6.5). The goal of the evader (v) is to
avoid a trajectory of this process from meeting with the set (6.5) on a whole
semi-infinite interval of time or if it is impossible to delay the moment of the
meeting as much as possible. The game is evolving in the time interval [0, T ],
where T is a moment when a trajectory of the process brings to a set (6.5)
M ∗ ;T > 0 such that z(T ) ∈ M ∗ or πz(T ) ∈ M,π is the orthogonal project
from Rn onto the subspace L [9].
In this game, we take the side of the pursuer. We will find sufficient
conditions on the parameters of the problem (6.1), (6.2), (6.5), ensuring the
game termination for a certain guaranteed time.
134 Method of Upper and Lower Resolving Functions

6.3 Scheme of the Method

Let Δ = {(t, s) : 0 ≤ s ≤ t < +∞}be the flat cone; ϕ(U, v) =
{ϕ(u, v) = u − v : u ∈ U } .
Consider the set-valued mappings W (t, s, v) = πK(t, s)ϕ(U, v) on the
set Δ × V, where K(t, s) is a matrix-valued function [31, 32].
For the practical finding of the fundamental matrix K(t, s) of systems
of differential-difference equations, it is more convenient to decompose this
function into a series.
Definition 6.1 [33–36]. For each s=1,2,. . . the time-delay exponential is
defined as follows
⎧
⎪ Θ, −∞ < t < −τ ;
⎪
⎨ I, −τ ≤ t < 0;
exp τ {B, t} = )2 s
⎪
⎪ I + B 1!t + B 2 (t−τ + ... + B k (t−(s−1)τ ) ,
⎩ 2! s!
(s − 1) τ ≤ t ≤ sτ.

Lemma 6.1 [35]. Let z(T ) be a continuous solution to the system (6.1) with
pure time-delay under the initial condition (6.2). Then,
50
z(T ) = exp τ {B, t}z 0 (−τ ) + −τ exp τ {B, t − τ − s}ż 0 (s)ds+
5t
+ 0 exp τ {B, t − τ − s}ϕ(u(s), v(s))ds.

Consider the set-valued mapping

W (t, s, v) = πexp τ (B, t)ϕ(U, v),


W (t, s) = W (t, s, v), 0 ≤ s ≤ t < +∞.
v∈V

Remark 6.1. For the linear process (6.1) (ϕ(u, v) = u − v) we have

W (t, s) = πexp τ {B, t}U ∗πexp τ {B, t}V,

∗ is a geometric subtraction of the sets (Minkowski’s difference) [38].

Let [28, 37]

domW = {(t, s) ∈ Δ : W (t, s) = ∅} .

Pontryagin Condition. domW = Δ.

It is further assumed that the Pontryagin condition is not satisfied.
 6.3 Scheme of the Method 135

Condition 6.1. The mapping W (t, s, v) is closed-valued on the direct product

of the cone Δ and compact V.
Let’s call the function γ(t, s), γ : Δ → L, a shift function. It’s almost
everywhere limited and measurable in t, integrated into s, s ∈ [0, t] for all
t ∈ [0, +∞).
We put

ξ(t) = ξ(t, z 0 (t), γ(t, ·)) = πexp τ {B, t}z 0 (−τ )+

50 5t
+ −τ πexp τ {B, t − τ − s}ż 0 (s)ds + 0 γ(t, s)ds.

Denote

R(t, s, v) = {α ≥ 0 : [πexp τ ϕ(u, v) − γ(t, s)] ∩ α [M − ξ(t)] = ∅} ,

 (6.6)
R : Δ × V → 2 , R(t, s) =
R+
R(t, s, v), (t, s) ∈ Δ.
v∈V

Denote the lower and upper resolving functions [28] first type

α∗ (t, s, v) = sup {α : α ∈ R(t, s, v)} ,

α∗ (t, s, v) = inf {α : α ∈ R(t, s, v)} .

Condition 6.2. The set-valued mapping R(t, s) has non-empty images on the
cone Δ.
Finally, denote the lower and upper resolving functions [28] second type:

α∗ (t, s) = sup {α : α ∈ R(t, s)} ,

α∗ (t, s) = inf {α : α ∈ R(t, s)} ,

and numerical functions
 t  t
α∗ (t) = α∗ (t, s)ds, α∗ (t) = α∗ (t, s)ds.
0 0

To the upper resolving function, which meaningfully means the maximum

gain of the pursuer at the moment s in the game on the interval [0, t] when
counteracting v, let’s put by the set
  t 
) *
T z 0 (·), γ(·, ·) = t ≥ 0 : inf α∗ (t, s, v(s)) ds ≥ 1 (6.7)
v(·)∈V (·) 0
136 Method of Upper and Lower Resolving Functions

and its element

) * 6 ) *7
t z 0 (·), γ(·, ·) = inf t : t ∈ T z 0 (·), γ(·, ·) .

If for some t, t > 0, α∗ (t, s, v) ≡ +∞, s ∈ [0, t], v ∈ V, then

in this case the value of the integral in relation (6.7) is naturally put equal
+∞,) and the corresponding
* inequality is performed automatically, and t ∈
T z 0 (·), γ(·, ·) . In the) case when *inequality in (6.6)
) 0does not hold
* for all
t > 0, we assume T z (·), γ(·, ·) = ∅, then, t z (·), γ(·, ·) = +∞,
0

respectively.
Theorem 6.1. Let the conflict-controlled process (6.1), (6.2), (6.5) satisfy
Conditions 6.1 and 6.2, M) = coM, and 0
* for a given function z (·) and a
shift functionγ(·, ·) T = T z (·), γ(·, ·) = ∅.
0

Then, the trajectory of the process (6.1), (6.2) can be under condition
α∗ (T ) < 1, brought by the pursuer (u) from z 0 (·) to the terminal set (6.3)
at the moment T by the control in the form of (6.4); and under the condition
α∗ (T ) > 1−in the counter-control class, with any permissible counteractions
of the evader.
Proof. Let v(·) be an arbitrary measurable function taking values from the
control domain V.
Suppose that α∗ (T, s, v) = +∞, s ∈ [0, T ], v ∈ V. We introduce the
controlling function
 t  T
h(t) = 1 − α∗ (T, s, v(s))ds − α∗ (T, s)ds, s ∈ [0, T ].
0 0
5T
If h(0) = 1 − 0 α∗ (T, s)ds = 1 − α∗ (T ) > 0,we have
 T
h(T ) = 1 − α∗ (T, s, v(s))ds ≤ 0.
0

Hence there exists a switching time t∗ , t∗ (v(·)), 0 < t∗ ≤ T, such that

h(t∗ ) = 0.
The whole process of approach is divided into time sections, such as
active [0, t∗ ) and passive [t∗ , T ], where t∗ is the moment of switching from
one law of choosing the counter-control of the pursuer to another, depending
on the prehistory of running away [9].
By the foregoing, we define the following law of choice of the control of
pursuer.
 6.3 Scheme of the Method 137

We consider the set-valued mapping

U1 (s, v) = {u ∈ U : πexp τ {B, T − τ − s}ϕ(u, v)−
− γ(T − τ − s) ∈ α∗ (T, s, v) [M − ξ(T )]} ,
U2 (s, v) = {u ∈ U : πexp τ {B, T − τ − s}ϕ(u, v)−
− γ(T − τ − s) = 0} , s ∈ [t∗ , T ]. (6.8)
It follows from the construction of the mappings U1 (s, v) and U2 (s, v)has
non-empty images. As stated in [28, 39–42] there are exist measurable
selections u1 (s, v) and u2 (s, v).
The pursuer’s control on the active interval [0, t∗ )is constructed in the
form
u1 (s) = u1 (s, v(s)).
On the passive interval [t∗ , T ] we set the control of pursuit in the following
form
u2 (s) = u2 (s, v(s)).
We will show that if the pursuer chooses control as described above, then the
trajectory of the process (6.1), (6.2) will reach the terminal set (6.5) at the
time T under arbitrary admissible controls of the evader.
By Lemma 6.1, the Cauchy formula for the system (6.1) under the
condition (6.2) implies the representation
 0
πz(T ) = πexp τ {B, T }z (−τ ) +
0
πexp τ {B, T − τ − s}ż 0 (s)ds+
−τ
 T
+ πexp τ {B, T − τ − s}ϕ(u(s), v(s))ds. (6.9)
0
5T
We add and subtract the value 0 γ(T, s)ds, applies (6.8), h(t∗ ) = 0, and
get the following:
 0
πz(T ) = πexp τ {B, T }z 0 (−τ ) + πexp τ {B, T − τ − s}ż 0 (s)ds+
−τ
 t∗
+ πexp τ {B, T − τ − s}ϕ(u1 (s), v(s))ds
0
 T
+ πexp τ {B, T − τ − s}ϕ(u2 (s), v(s))ds±
t∗
 T
± γ(T, s)ds ∈ πexp τ {B, T }z 0 (−τ )
0
138 Method of Upper and Lower Resolving Functions

 t∗
+ α∗ (T, s, v(s))[M − ξ(T )]ds+
0
 T  T
∗
+ α (T, s)[M − ξ(T )]ds + γ(T, s)ds =
t∗ 0
  t∗  T 
∗ ∗
= ξ(T ) 1 − α (T, s, v(s))ds − α (T, s)ds +
0 t∗
 t∗  T
∗
+ α (T, s, v(s))M ds + α∗ (T, s)M ds =
0 t∗
 T  T 
∗ ∗
= α (T, s)ds + α (T, s)ds M = M.
t∗ t∗

From (6.6) we have that the case α∗ (T, s, v) = +∞ for some s ∈ [0, T ], v ∈
V, is possible only if 0 ∈ M − ξ(T ), 0 ∈ exp τ {B, T }ϕ(U, v) − γ(T, s), and
in this case it is obvious that R(T, s, v) = [0, +∞), s ∈ [0, T ], v ∈ V, and
α∗ (T, s) = 0, s ∈ [0, T ].
This makes it possible to choose as the resolving functions at those points
s ∈ [0, T ] where α∗ (T, s, v(s)) = +∞, an arbitrary finite measurable
function that takes values on the interval [0, +∞) with only one condition
that the final function on the interval [0, T] provides relation h(t∗ ) = 0 for a
certain switching moment t∗ , t∗ ∈ [0, T ].
Thus, the construction of control in the active and the passive time
sections is reduced to the previous case.

6.4 Conclusion
The concept of upper and lower resolving functions of two types for
differential-difference games of approach with pure time delay is introduced
at the first time. Sufficient conditions are obtained for the trajectory of the
process (6.1), (6.2) to converge with a given cylindrical terminal set (6.5) in
the case when the Pontryagin condition does not hold.
In the future, it is planned to expand the proposed scheme for processes
that are described by systems of differential-difference equations with several
delays based on the use of the Cauchy formulas [43–45].
It is planned to apply the proposed scheme to specific examples and
to realize the visualization of the trajectory of the objects on the plane
[46, 47].
 References 139

We also plan to shift the proposed scheme of the method of revolving

functions for group pursuit problems [48, 49] and nonstationary pursuit
[50–53].

References
[1] Kuntsevich, V.M., et al (Eds). Control Systems: Theory and Applica-
tions. Series in Automation, Control, and Robotics, River Publishers,
Gistrup, Delft, 2018.
[2] Kondratenko, Y.P., Kuntsevich, V.M., Chikrii, A.A., Gubarev, V.F.
(Eds). Advanced Control Systems: Theory and Applications. Series in
Automation, Control, and Robotics, River Publishers, Gistrup, 2021.
[3] Kondratenko, Y.P., Chikrii, A.A., Gubarev, V.F., Kacprzyk, J. (Eds).
Advanced Control Techniques in Complex Engineering Systems: Theory
and Applications. Dedicated to Professor Vsevolod M. Kuntsevich.
Studies in Systems, Decision, and Control, Vol. 203. Cham: Springer
Nature Switzerland AG, 2019.
[4] R. Isaacs. Differential Games: A Mathematical Theory with Applications
to Warfare and Pursuit, Control and Optimization. John Wiley & Sons
Inc, New York, 1965.
[5] L.S. Pontryagin. Selected scientific works. Nauka, Moscow, 2, 1988.
[6] M.S. Nikol’skii. Pontryagin’s First Direct Method in Differential
Games. Mosk. Gos. Univ., Moscow, 1984 (in Russian).
[7] N.N. Krasovskii, A.I. Subbotin. Game-theoretical control problems.
Springer-Verlag, New York, 1988.
[8] V.N. Ushakov. Krasovskii’s Unification Method and the Stability Defect
of Sets in a Game Problem of Approach on a Finite Time Interval. In:
Kondratenko Y., Chikrii A., Gubarev V., Kacprzyk J. (eds) Advanced
Control Techniques in Complex Engineering Systems: Theory and
Applications. Studies in Systems, Decision, and Control, vol 203.
Springer, Cham, 2019. https://doi.org/10.1007/978-3-030-21927-7_5.
[9] A. Chikrii. Conflict-Controlled Processes. Springer Science &Business
Media, 2013.
[10] A.A. Chikrii, I.S Rappoport, K.A. Chikrii. Multivalued Mappings and
their Selectors in the Theory of Conflict-Controlled Processes. Cyber-
netics and Systems Analysis, 43(5):719–730, 2007.
[11] A.A. Chikrii, R. Petryshyn, I. Cherevko., Y. Bigun. Method of Resolving
Functions in the Theory of Conflict—Controlled Processes. In: Kon-
dratenko Y., Chikrii A., Gubarev V., Kacprzyk J. (eds) Advanced Control
140 Method of Upper and Lower Resolving Functions

Techniques in Complex Engineering Systems: Theory and Applications.

Studies in Systems, Decision, and Control, vol 203. Springer, Cham,
2019. https://doi.org/10.1007/978-3-030-21927-7_1.
[12] J. Albus, A. Meystel, A.A. Chikrij, A.A. Belousov, A.I. Kozlov. Ana-
lytical Method for Solution of the Game Problem of Self Landing for
Moving Objects. Cybernetics and Systems Analysis, 37(1):75–91, 2001.
[13] A.A. Chikrii, S.F. Kalashnikova. Pursuit of a Group of Evaders by a
Single Controlled Object. Cybernetics, 23(4):437–445, 1987.
[14] A.A. Chikrii, I.S. Rappoport. Systems Analysis Method of Resolving
Functions in the Theory of Conflict-Controlled Processes. Cybernetics
and Systems Analysis, 48(4):512–531, 2012.
[15] A.A. Chikrii, G.T. Chikrii. Matrix Resolving Functions in Game Prob-
lems of Dynamics. Proceedings of the Steklov Institute of Mathematics
268(SUPPL. 1), 2010.
[16] L.V. Baranovska. Quasi-Linear Differential-Difference Game of
Approach. Understanding Complex Systems, 2019, pp. 505–524.
https://doi.org/10.1007/978-3-319-96755-4_26.
[17] L.V. Baranovska. On Quasilinear Differential-Difference Games of
Approach. Journal of Automation and Information Sciences, 49(8):53–
67, 2017. https://doi.org/10.1615/JAutomatInfScien.v49.i8.40.
[18] Lesia V. Baranovska. Pursuit differential-difference games with pure
time-lag. Discrete and Continuous Dynamical Systems – Series B
24(3):1021–1031, 2019. https://doi.org/10.3934/dcdsb.2019004.
[19] L.V. Baranovska. Group Pursuit Differential Games with Pure
Time-Lag. In: Sadovnichiy, V., Zgurovsky, M. (eds.) Contemporary
Approaches and Methods in Fundamental Mathematics and Mechanics.
Understanding Complex Systems, pp. 475–488. Springer, Cham, 2021.
https://doi.org/10.1007/978-3-030-50302-4_23.
[20] G.G. Baranovskaya, L.V. Baranovskaya. Group Pursuit in Quasilinear
Differential-Difference Games. Journal of Automation and Information
Sciences, 29(1):55–62, 1997. https://doi.org/10.1615/JAutomatInfScien
.v29.i1.70.
[21] L.V. Baranovskaya, A.A. Chikrij, Al.A. Chikrij. Inverse Minkowski
functionals in a nonstationary problem of group. Izvestiya Akademii
Nauk. Teoriya i Sistemy Upravleniya, (1):109–114, 1997.
[22] L.V. Baranovskaya, A.A. Chikrij, Al.A. Chikrij. Inverse Minkowski
functionals in a nonstationary problem of group. Journal of Computer
and Systems Sciences International, 36(1):101–106, 1997.
 References 141

[23] L.V. Baranovskaya. A method of resolving functions for one

class of pursuit problems. Eastern-European Journal of Enterprise
Technologies, vol.2, 4(74):4–8, 2015. https://doi.org/10.15587/1729-
4061.2015.39355.
[24] L.V. Baranovska. Method of resolving functions for the differential-
difference pursuit game for different-inertia objects. Studies in Systems,
Decision and Control, 69:159–176, 2016. https://doi.org/10.1007/978-
3-319-40673-2_7.
[25] A.A. Chikrij, L.V. Baranovskaya, Al.A. Chikrij. The game problem of
approach under the condition of failure of controlling devices. Problemy
Upravleniya i Informatiki (Avtomatika), (4):5–13, 1997.
[26] L.V. Baranovskaya, Al.A. Chikrii. Game Problems for a Class of
Hereditary Systems. Journal of Automation and Information Sciences,
29(2):87–97, 1997. https://doi.org/10.1615/JAutomatInfScien.v29.i2-
3.120.
[27] A.A. Chikriy, L.V. Baranovskaya, Al.A. Chikriy. An Approach Game
Problem under the Failure of Controlling Devices. Journal of Automa-
tion and Information Sciences, 35(5):1–8, 2000. https://doi.org/10.161
5/JAutomatInfScien.v32.i5.10.
[28] A.A. Chikrii, V.K. Chikriy. Image Structure of Multivalued Mappings
in Game Problems of Motion Control. Journal of Automation and Infor-
mation Science, 48(3):20–35, 2016. https://doi.org/10.1615/jautomatin
fscien.v48.i3.30.
[29] A.A. Chikrii. An analytical method in dynamic pursuit games, Pro-
ceedings of the Steklov Institute of Mathematics, vol. 271, no. 1, 2010,
pp. 85.
[30] O. Hayek. Pursuit Games (Academic, New York, 1975), Math. Sci. Eng.
120.
[31] L.E. Elsgolts, S.B. Notkin. Differential Equations with Deviating Argu-
ment. Nauka, Moscow, 1971.
[32] R. Bellman, K. Cooke. Differential-Difference Equations, Academic,
Cambridge, 1963.
[33] D.Ya. Khusainov, D.D. Benditkis, J. Diblik. Weak delay in system with
an aftereffect. Funct. Differ. Equ. 9(3-4):385-404, 2002.
[34] J. Diblik, B. Morvkov, D. Khusainov, A. Kukharenko. Delayed expo-
nential functions and their application to representations of solutions
of linear equations with constant coefficients and with single delay. In:
Proceedings of 2nd International Conference on Mathematical Models
for Engineering Science, 2011.
142 Method of Upper and Lower Resolving Functions

[35] D.Ya. Khusainov, J. Diblik, M. Ruzhichkova. Linear dynamical sys-

tems with aftereffect. In: Representation of Decisions, Stability, Control,
Stabilization. GP Inform-Analytics Agency, Kyiv, 2015.
[36] J. Diblik, D. Khusainov. Representation of solutions of linear dis-
crete systems with constant coefficients and pure delay. Advances in
Difference Equations, 2006. https://doi.org/10.1155/ADE/2006/80825.
[37] R.T. Rockafellar. Convex Analysis. Princeton University Press, Prince-
ton, 1970.
[38] B.N. Phenichnyi. Simple Pursuit by Several Objects. Kibernetika,
3:145-146, 1976.
[39] J.-P. Aubin, I. Ekeland. Applied Nonlinear Analysis, Wiley, New York,
1984.
[40] J.-P. Aubin, He. Francowska. Set-Valued Analysis. Birkhause, Boston,
1990.
[41] R.J. Aumann. Integrals of set-valued functions. J. Math. Anal. Appl.,
12:1–12, 1965.
[42] A.A. Chikrii. Multivalued mappings and their selections in game control
problems. Journal of Automation and Information Science, 27(1):27–38,
1995.
[43] M. Medved’, M. Pospíŝil. Sufficient conditions for the asymptotic sta-
bility of nonlinear multidelay differential equations with linear parts
defined by pairwise permutable matrices. Nonlinear Anal., 75:3348–
3363, 2012.
[44] M. Pospíŝil. Representation of solutions of systems of linear differential
equations with multiple delays and nonpermutable variable coefficients.
Mathematical modeling and analysis, 25(2):303–322, 2020.
[45] M. Pospíŝil, F. Jaroŝ. On the representation of solutions of delayed
differential equations via Laplace transform. Electronic Journal of
Qualitative Theory of Differential Equations, 117:1–13, 2016.
[46] L. Baranovska, D. Hyriavets, K. Dovzhanytsia, V. Mukhin. Visual-
ization of pursuit differential game on a plane. Selected Papers of
the XX International Scientific and Practical Conference “Information
Technologies and Security” (ITS 2020), 2020, pp. 99–113.
[47] A.A. Chikrii, L.A. Sobolenko, S.F. Kalashnikova. A numerical method
for the solution of the pursuit-and-evasion problem. Cybernetics,
24(1):53–59, 1988.
[48] M. Pittsyk, A.A. Chikrii. On group pursuit problem. Journal of Applied
Mathematics and Mechanics, 46(5):584–589, 1982.
 References 143

[49] A.A. Chikrii, P.V. Prokopovich. Simple pursuit of one evader by a group.
Cybernetics and Systems Analysis, 28(3):438–444, 1992.
[50] Al.A. Chikrii. On nonstationary game problem of motion control. Jour-
nal of Automation and Information Sciences, 47(11):74-83, 2015.
https://doi.org/10.1615/JAutomatInfScien.v47.i11.60.
[51] I.I. Kryvonos, Al.A. Chikrii, K.A. Chikrii. On an approach scheme in
nonstationary game problems. Journal of Automation and Information
Sciences, 45(80):32–40, 2013. https://doi.org/10.1615/JAutomatInfSci
en.v45.i8.40.
[52] V.A. Pepelyaev, Al.A. Chikrii. On the game dynamics problems for non-
stationary controlled processes. Journal of Automation and Information
Sciences, 49(3):13–23, 2017. https://doi.org/10.1615/JAutomatInfScien
.v49.i3.30.
[53] Y.N. Onopchuk, Al.A. Chikrii. The analytical method to solve nonsta-
tionary differential games of pursuit. Cybernetics and Systems Analy-
sis, 49(4):603–615, 2013. https://doi.org/10.1007/s10559-013-9547-7.
 7
Adaptive Method for the Variational
Inequality Problem Over the Set of
Solutions of the Equilibrium Problem

Y. Vedel, S. Denisov, and V. Semenov

Faculty of Computer Science and Cybernetics, Taras Shevchenko National

University of Kyiv, Kyiv, Ukraine
E-mail: {yana.vedel, denisov.univ, semenov.volodya}@gmail.com

Abstract
In this chapter we consider a two-level variational problem: a variational
inequality problem over the set of solutions of the equilibrium problem. An
example of such a problem is the search for the normal Nash equilibrium.
To solve this problem, we propose a new iterative proximal algorithm that
combines the ideas of a two-stage proximal point method, adaptability, and
scheme of iterative regularization. In contrast to the previously applied rules
for choosing the step size, the proposed algorithm does not calculate any
values of the bifunction at additional points; it does not require knowledge
of information about the Lipschitz constants of the bifunction, the Lipschitz
constant and strong monotonicity constant of the operator. For monotone Lip-
schitzian bifunctions and strongly monotone Lipschitz continuous operators,
a theorem on the strong convergence of the algorithm is proved. It is shown
that the proposed algorithm is applicable to monotone two-level variational
inequalities in real Hilbert spaces.

Keywords: Two-level problem, variational inequality problem, equilibrium

problem, strong convergence, two-stage proximal method, scheme of iterative
regularization.

145
146 Adaptive Method for the Variational Inequality Problem Over the Set

7.1 Introduction
In numerical optimization, and the theory of ill-posed (non-correct) problems,
the following technique for solving problems with a non-unique solution
is widespread [1–4]. The problem is associated with a family of perturbed
auxiliary problems that are uniquely and correctly (well-posed) solvable.
A particular solution to the original problem is obtained as the limit of
solutions to perturbed auxiliary problems with a decrease in perturbations.
The solutions found in this way satisfy certain additional conditions. For
example, the minimality of the norm of the normal solution to an optimization
problem obtained by the Tikhonov regularization algorithm. On the other
hand, in the operation research, optimization problems arise according to
sequentially specified criteria (lexicographic, sequential, or multi-stage opti-
mization) [5, 6]. Note that the ideas of the Tikhonov regularization method
permeate many studies in the field of automatic control [7–10].
Two-level variational inequalities arose as a natural generalization of well
studied lexicographic optimization problems with two criteria, as well as in
the analysis of ordinary optimization problems with functional constraints
in the form of a monotone variational inequality. As an independent math-
ematical object, two-level variational inequalities began to be considered in
[11]. The papers [12, 13] are devoted to the solvability of more general n-
level variational inequalities. By now, approximate methods of one-stage for
solution of all these problems are known, which are ideologically close to the
methods of penalty and regularization [6, 14].
One of the popular directions of contemporary applied nonlinear analysis
is the study of equilibrium problems (equilibrium programming problems)
[15–23], to which mathematical programming problems, variational inequal-
ities and many game theory problems can be reduced. Recently, algorithms
for solving variational inequalities and equilibrium problems have become
popular in the ML community. On their basis it is possible to design effective
learning methods for GANs [24–26].
The Chapter deals with a two-level problem: variational inequality prob-
lem on the set of solutions to the equilibrium programming problem. A
similar problem was considered in [22], where a strongly converging algo-
rithm was proposed that used the operation of calculating the value of the
resolvent of a bifunction. The latter significantly increased the complexity
of the algorithm. Below, to solve this problem, an adaptive version of the
previously studied algorithm [23] based on the two-stage proximal point
method [20, 21] and the idea of scheme of iterative regularization [1, 27]
 7.2 Problem Formulation 147

will be proposed. The proposed algorithm does not calculate the values of
the bifunction at any additional points; it does not require knowledge of
information about the Lipschitzian constants of the bifunction, the constants
of Lipschitz and strong monotonicity of the operator.
For monotone bifunctions of Lipschitzian type and strongly monotone
Lipschitz continuous operators, a theorem on the strong convergence of the
algorithm is proved. It is shown that the proposed algorithm is applicable to
monotone two-level variational inequalities in real Hilbert spaces.
The results obtained are useful in the theory of optimal control. Indeed,
the problem of optimal control of the system can be formulated as a vari-
ational inequality problem. In the case of non-uniqueness of the optimal
control, an additional criterion can be introduced for the regularization of
the problem and a two-level variational in equality can be obtained. And to
solve the obtained problem, we use the algorithm proposed and studied in this
Chapter.

7.2 Problem Formulation

Everywhere below H is real infinite-dimensional Hilbert space with inner
product (·, ·) and induced norm ·. For an operator T : H → H, set D ⊆ H,
and bifunction G : H × H → R we denote V I (T, D) and EP (G, D) sets
{ξ ∈ D : (T ξ, η − ξ) ≥ 0 ∀η ∈ D} ,
{ξ ∈ D : G (ξ, η) ≥ 0 ∀η ∈ D} ,
respectively.
For closed convex set C ⊆ H consider two-level problem:
find x ∈ V I (F, EP (G, C)) . (7.1)
Assume the following conditions hold [23]:
C1) G (x, x) = 0 ∀x ∈ C;
C2) G (x, y) + G (y, x) ≤ 0 ∀x,y ∈ C(monotonicity);
C3) ∀x ∈ C functional G (x, ·) is closed and convex on C;
C4) ∀y ∈ C functional −G (·, y) is weakly closed on C;
C5) for all x,y,z ∈ C holds
G (x, y) ≤ G (x, z) + G (z, y) + a x − z2 + b z − y2 ,
where a,b are positive constants (Lipschitz continuity);
148 Adaptive Method for the Variational Inequality Problem Over the Set

C6) EP (G, C) = ∅;
C7) F : C → H–μ-strongly monotone and L-Lipschitz continuous operator.
Remark 7.1. Condition C5 is provided by Giandomenico Mastroeni in [16].
For example, bifunction

G (x, y) = (T x, y − x)

with Lipschitz continuous operator T : C → H satisfies C5 with a = Lε

2 and
L
b = 2ε where L > 0 is Lipschitzian constant of operator T , ε > 0.
With given conditions the set EP (G, C) is closed and convex and
problem (7.1) has unique solution x̄ ∈ C [17].
Remark 7.2. The most known case of (7.1) is search of normal solution of
variational inequality (with F x = 2x, G (x, y) = (T x, y − x) where T :
C → H):

x2 → min, x ∈ C : (T x, y − x) ≥ 0 ∀ y ∈ C.

Let us remind one important definition. Let ϕ : H → R ∪ {+∞} be proper

convex closed functional. Proximal operator associated with functional ϕ is
operator  
proxϕ x = arg min ϕ (η) + 12 η − x2 .
η∈dom ϕ

This operator is firmly non-expansive and

ϕ (η) − ϕ (z) + (z − x, η − z) ≥ 0

for all η ∈ dom ϕ if and only if z = proxϕ x [28].

Below we provide strongly converging iterative algorithm for solving the
two-level variational problem (7.1) that does not require knowledge of con-
stants from conditions C5, C7. First, we approximate the two-level problem
(7.1) by one-level and more regular problem of equilibrium programming.

7.3 Tikhonov–Browder Approximation

Consider auxiliary equilibrium problem depending on small parameter ε > 0:

find x ∈ C : G (x, y) + ε (F x, y − x) ≥ 0 ∀ y ∈ C. (7.2)

Following Bakushinskii [1], we call the equilibrium problem (7.2) the

Tikhonov–Browder approximation of two-level problem (7.1).
 7.3 Tikhonov–Browder Approximation 149

Remark 7.3. For solving incorrect extremum problems that type of approxi-
mation was proposed by Tikhonov for building regularization methods. Later
Felix Browder [2, 3] applied similar regularization technique.
From results of [17] follows the existence and uniqueness of solution
xε ∈ C of problem (7.2) for all ε > 0. Auxiliary elements xε ∈ C have
several properties important for subsequent constructions.
Lemma 7.1 ([23]). The following inequalities take place:
 
1) xε  ≤ μ1 1 + μ1 L F x̄ ∀ε > 0;
 
2) xε − xδ  ≤ |ε − δ| εμ
1
1 + L μ1 F x̄ ∀ε, δ > 0.
Proof. Let ε > 0. For xε ∈ C and an arbitrary element w ∈ EP (G, C), we
have

G (xε , w) + ε (F xε , w − xε ) ≥ 0 and G (w, xε ) ≥ 0.

Adding the inequalities and using the monotonicity of the bifunction G, we

obtain
(F xε , w − xε ) ≥ 0,
that is,
(F w, w − xε ) ≥ (F xε − F w, xε − w) .
The strong monotonicity of the operator F and the Cauchy–Schwarz inequal-
ity give
μ w − xε  ≤ F xε  ,
whence 1) follows.
Let us prove 2). Let xε ∈ C and xδ ∈ C be solutions of problem (7.2) with
ε > 0 and δ > 0, respectively. We have

G (xε , xδ ) + ε (F xε , xδ − xε ) ≥ 0 and G (xδ , xε ) + δ (F xδ , xε − xδ ) ≥ 0.

Adding the inequalities and using the monotonicity of the bifunction G,

we obtain
ε (F xε , xδ − xε ) + δ (F xδ , xε − xδ ) ≥ 0.
Let us rewrite the last inequality us

(δ − ε) (F xδ , xε − xδ ) ≥ ε (F xε − F xδ , xε − xδ ) .
150 Adaptive Method for the Variational Inequality Problem Over the Set

Using the strong monotonicity of the operator F , we obtain

|δ − ε| F xδ  xε − xδ  ≥ εμ xε − xδ 2 ,

that is,
|δ − ε|
xε − xδ  ≤ F xδ  .
εμ
Let us estimate from above the norm of F xδ
 
L
F xδ  ≤ F x̄ + F xδ − F x̄ ≤ F x̄ + L xδ − x̄ ≤ 1+ F x̄ .
μ
Using the last inequality in the estimate for xε − xδ , we obtain 2). 
Lemma 7.2 ([23]). Assume that conditions C1–C4 and C6, C7 are satisfied.
Then
lim xε − x̄ = 0.
ε→0
Proof. By first inequality of Lemma 7.1, from {xε }ε>0 we can extract a
sequence (xεn ) (εn → 0) that weakly converges to an element v ∈ C. Using
the weak upper semicontinuity of the functional G (·, y), we pass to the limit
in the inequality

G (xεn , y) + εn (F xεn , y − xεn ) ≥ 0 ∀y ∈ C.

We get
G (v, y) ≥ 0 ∀y ∈ C,
that is, v ∈ EP (G, C). And passing to the limit in inequality

(F w, w − xεn ) ≥ (F xεn , w − xεn ) ≥ 0 ∀w ∈ EP (G, C) ,

we obtain
(F w, w − v) ≥ 0 ∀w ∈ EP (G, C) ,
that is, v = x̄. Let us show that

lim xε − x̄ = 0.

ε→0

This follows from inequality

(F x̄, x̄ − xεn ) ≥ (F xεn − F x̄, xεn − x̄) ≥ μ xεn − x̄2 .

From the uniqueness of the element x̄ ∈ H we obtain limε→0 xε − x̄ = 0.
 7.4 Algorithm 151

7.4 Algorithm
Following the well-known scheme of two-step proximal point method [20,
21] for solving problem (7.1) in recent work [23] the following method was
introduced: ⎧
⎨zn = xn − αn λn F xn ,
yn = proxλn G(yn−1 ,·) zn , (7.3)
⎩
xn+1 = proxλn G(yn ,·) zn ,
8 9  
where λn were defined considering λn ∈ λ, λ̄ ⊆ 0, 2(2a+b) 1
and positive
sequence (αn ) was such that
∞
! αn+1 − αn
lim αn = 0, αn = +∞, lim = 0.
n→∞ n→∞ αn2
n=1

For getting rid of explicit usage of information about values of constants a

and b with defined limits for λn in (7.3) we consider the following algorithm
with adaptiveness in choice of parameters λn .
Algorithm 7.1. ) *
Initialization. Choose elements x1 , y0 ∈ C, and τ ∈ 0, 13 , λ1 ∈ (0, +∞).
Let n = 1.
Step 1. Calculate
zn = xn − αn λn F xn .
Step 2. Calculate
yn = proxλn G(yn−1 ,·) zn .
Step 3. Calculate
xn+1 = proxλn G(yn ,·) zn .
Step 4. Calculate
⎧
⎨λn , if G (yn−1 , xn+1 ) − G (yn−1 , yn )−G (yn , xn+1 ) ≤ 0,
λn+1 = 
⎩min λn , τ yn−1 −yn 2 +yn −xn+1 2
2 (G(yn−1 ,xn+1 )−G(yn−1 ,yn )−G(yn ,xn+1 )) , otherwise.

Let n := n + 1 and go to step 1.

In the proposed iterative algorithm, parameter λn+1 depends on locations
of points yn−1 , yn , xn+1 and values of G (yn−1 , xn+1 ), G (yn−1 , yn ), and
G (yn , xn+1 ). We do not use any information about constants a and b from
152 Adaptive Method for the Variational Inequality Problem Over the Set

the inequality. Obviously,

the sequence (λn) is decreasing. Also, it is lower
bounded by min λ1 , τ 2 (max {a, b})−1 . We have
−1

G (yn−1 , xn+1 ) − G (yn−1 , yn ) − G (yn , xn+1 )

 
≤ max {a, b} yn−1 − yn 2 + yn − xn+1 2 .
Regarding positive parameters αn we assume that the next conditions are
satisfied:
D1) lim αn = 0;

n→∞
D2) ∞ n=1 αn = +∞;
−αn
D3) lim αn+1
α2
= 0.
n→∞ n

Remark 7.4. For (αn ) we can use αn = 1

np , p ∈ (0, 1).

7.5 Proof of Algorithm Convergence

We provide proof of convergence with the following simple scheme. Let xαn
be a solution of auxiliary problem (7.2) for ε = αn . Since

xn − x̄ ≤ xn − xαn  + xαn − x̄ ,

lim xαn − x̄ = 0

n→∞

it is enough to show that the generated by Algorithm 7.1 sequence (xn ) has
the next property
lim xn − xαn  = 0.
n→∞
Let us remind the well-known elementary lemma about recurrent numerical
inequalities.
Lemma 7.3 ([1]). Let the non-negative numerical sequence (ψn ) satisfies the
recurrent inequality

ψn+1 ≤ (1 − αn ) ψn + βn for all n ∈ N

where numerical sequences (αn ) and (βn ) have the following conditions

αn ∈ (0, 1), αn = +∞, lim αβnn ≤ 0. Then lim ψn = 0.
n→∞ n→∞
Let us prove an important inequality for generated sequences (xn ), (yn )
and auxiliary elements xαn .
 7.5 Proof of Algorithm Convergence 153

Lemma 7.4. For sequences generated by Algorithm 7.1 (xn ), (yn ) and
auxiliary elements xαn the inequality holds
 
2 2 λn
xn+1 − xαn  ≤ (1 − αn λn μ) xn − xαn  − 1 − τ xn+1 − yn 2
λn+1
 
λn 2 −1
− 1 − 2τ − α n λn L μ yn − xn 2
λn+1
λn
+ 2τ xn − yn−1 2 . (7.4)
λn+1
Proof. We have
xn+1 − xαn 2 = xn − xαn 2 − xn − xn+1 2
+ 2 (xn+1 − xn , xn+1 − xαn ) = xn − xαn 2 − xn − yn 2
−yn − xn+1 2 − 2 (xn − yn , yn − xn+1 )
+ 2 (xn+1 − xn , xn+1 − xαn ) . (7.5)
From the definition of points xn+1 and yn it follows that
λn G (yn , xαn ) − λn G (yn , xn+1 )
≥ (xn+1 − xn + αn λn F xn , xn+1 − xαn ) , (7.6)
λn G (yn−1 , xn+1 ) − λn G (yn−1 , yn )
≥ − (xn − αn λn F xn − yn , yn − xn+1 ) . (7.7)
Using inequalities (7.6), (7.7) for estimating inner products in (7.5),
we get
xn+1 − xαn 2 ≤ xn − xαn 2 − xn − yn 2 − yn − xn+1 2
+ 2λn (G (yn , xαn ) − G (yn , xn+1 ) + G (yn−1 , xn+1 ) − G (yn−1 , yn ))
+ 2αn λn (F xn , xαn − yn ) . (7.8)
From the rule of calculation of λn+1 it follows the inequality
G (yn−1 , xn+1 ) − G (yn−1 , yn ) − G (yn , xn+1 )
1  
≤τ yn−1 − yn 2 + xn+1 − yn 2 . (7.9)
2λn+1
For estimating an expression G (yn−1 , xn+1 ) − G (yn−1 , yn ) −
G (yn , xn+1 ) in (7.8) we use (7.9). We obtain
xn+1 − xαn 2 ≤ xn − xαn 2 − xn − yn 2 − yn − xn+1 2
154 Adaptive Method for the Variational Inequality Problem Over the Set

λn λn
+τ yn−1 − yn 2 + τ xn+1 − yn 2
λn+1 λn+1
+2λn G (yn , xαn ) + 2αn λn (F xn , xαn − yn ) .
We estimate yn−1 − yn 2 in the following way

yn−1 − yn 2 ≤ 2 yn−1 − xn 2 + 2 xn − yn 2 .

We have

xn+1 − xαn 2 ≤ xn − xαn 2 − xn − yn 2 − yn − xn+1 2

λn λn λn
+ 2τ yn−1 − xn 2 + 2τ xn − yn 2 + τ xn+1 − yn 2
λn+1 λn+1 λn+1
+ 2λn G (yn , xαn ) + 2αn λn (F xn , xαn − yn ) . (7.10)

From monotonicity of bifunction G it follows G (yn , xαn ) ≤ G (xαn , yn )

hence

G (yn , xαn )−αn (F xαn , yn − xαn ) ≤ −G (xαn , yn )−αn (F xαn , yn − xαn ) .

Since
G (xαn , yn ) + αn (F xαn , yn − xαn ) ≥ 0
then
G (yn , xαn ) ≤ αn (F xαn , yn − xαn ) .
Taking into account the last estimation in (7.10) we obtain

xn+1 − xαn 2 ≤ xn − xαn 2 − xn − yn 2 − yn − xn+1 2

λn λn λn
+ 2τ yn−1 − xn 2 + 2τ xn − yn 2 + τ xn+1 − yn 2
λn+1 λn+1 λn+1
+ 2αn λn (F xn − F xαn , xαn − yn ) . (7.11)

Let us estimate the upper bound for (F xn − F xαn , xαn − yn ). We have

(F xn − F xαn , xαn − yn ) = (F xn − F xαn , xαn − xn )

+ (F xn − F xαn , xn − yn ) ≤ −μ xαn − xn 2
μ
+ L xn − xαn  xn − yn  ≤ −μ xαn − xn 2 + xn − xαn 2
2
L2 μ L2
+ xn − yn 2 = − xn − xαn 2 + xn − yn 2 . (7.12)
2μ 2 2μ
 7.5 Proof of Algorithm Convergence 155

From inequalities (7.11) and (7.12) we obtain

) *
xn+1 − xαn 2 ≤ (1 − αn λn μ) xn − xαn 2 − 1 − τ λn λ−1
n+1
) *
× xn+1 − yn 2 − 1 − 2τ λn λ−1 2 −1
n+1 − αn λn L μ

× yn − xn 2 + 2τ λn λ−1 2
n+1 xn − yn−1  ,
Q.E.D. 
Let us prove the estimation from which it follows the convergence to 0 of
sequences (xn − xαn ) and (xn − yn−1 ).
Lemma 7.5. For sequences (xn ), (yn ) generated by Algorithm 7.1 and
auxiliary elements xαn for large n the inequality is hold

 2 2τ λn+1 λ−1
xn+1 − xα  + n+2
xn+1 − yn 2
n+1
1 − αn+1 λn+1 μ
 1 2
αn λ n μ 2 2τ λn λ−1
n+1 2
≤ 1− xn − xαn  + xn − yn−1 
2 1 − α n λn μ
2M (αn+1 − αn )2
+ , (7.13)
λn μ αn3
) *
where M = μ−1 1 + Lμ−1 F x̄.
Proof. We have
 2  2
xn+1 − xαn 2 = xn+1 − xαn+1  + xαn+1 − xαn 
) *  2  2
+ 2 xn+1 − xαn+1 , xαn+1 − xαn ≥ xn+1 − xαn+1  + xαn+1 − xαn 
    2
− 2 xn+1 − xαn+1  xαn+1 − xαn  ≥ (1 − ε) xn+1 − xαn+1 
ε−1 xα
2
 ,
+ n+1 − xαn (7.14)
ε
where ε > 0. Let in (7.14) ε = 12 αn λn μ. We get

2 − α n λn μ 
xn+1 − xα 2

xn+1 − xαn 2 ≥ n+1
2
2 − α n λn μ 
xα
2
 .
− n+1 − xαn (7.15)
α n λn μ
156 Adaptive Method for the Variational Inequality Problem Over the Set

With rules for values of algorithm’s parameters αn , λn for large numbers n

we have 1 − αn λn μ > 0. Taking into consideration the second inequality of
Lemma 7.1 from (7.15) we obtain
 
2 αn λ n μ xn+1 − xα 2

xn+1 − xαn  ≥ 1 − n+1
2
(2 − αn λn μ) ) *
− (αn+1 − αn )2 αn−2 μ−1 1 + Lμ−1 F x̄ ,
α n λn μ
(7.16)
for all numbers n ≥ n0 . Using (7.16) in (7.4) we get (for numbers n ≥ n0 )
2 − αn λn μ  
xn+1 − xα 2 ≤ (1 − αn λn μ) xn − xαn 2
n+1
) 2 * ) *
− 1 − τ λn λ−1 2 −1 2 −1
n+1 xn+1 − yn  − 1 − 2τ λn λn+1 − αn λn L μ
(2 − αn λn μ) (αn+1 − αn )2
× yn − xn 2 +2τ λn λ−1 2
n+1 xn − yn−1  + M,
α n λn μ αn2
) *
where M = μ−1 1 + Lμ−1 F x̄. Hence it follows
) *
 2 2 − 2αn λn μ 2 1 − τ λn λ−1
xn+1 − xα  ≤ 2
xn − xαn  − n+1
n+1
2 − α n λn μ 2 − α n λn μ
) *
2 2 1 − 2τ λn λ−1 2 −1
n+1 − αn λn L μ
× xn+1 − yn  − yn − xn 2
2 − α n λn μ
4τ λn λ−1
n+1 2M (αn+1 − αn )2
+ xn − yn−1 2 + . (7.17)
2 − α n λn μ λn μ αn3
Regrouping the variables in (7.17) we obtain
 2
xn+1 − xα  + 2τ λn+1 1
xn+1 − yn 2
n+1
λn+2 1 − αn+1 λn+1 μ
 
2 − 2αn λn μ 2 λn 1 2
≤ xn − xαn  + 2τ xn − yn−1 
2 − α n λn μ λn+1 1 − αn λn μ
) *
2 1 − 2τ λn λ−1 2 −1
n+1 − αn λn L μ
− yn − xn 2
2 − α n λn μ
1 2
1 − τ λn λ−1n+1 2τ λn+1 λ−1
n+2
− − xn+1 − yn 2 .
1 − 12 αn λn μ 1 − αn+1 λn+1 μ
 7.5 Proof of Algorithm Convergence 157

2M (αn+1 − αn )2
+ . (7.18)
λn μ αn3
) *
Since τ ∈ 0, 13 there exists lim λn > 0 and lim αn = 0, then starting
n→∞ n→∞
from some number n1 the inequalities

1 − 2τ λλn − αn λn L2 μ−1
n+1
> 0,
2 − α n λn μ
1 − τ λλn 2τ λλn+1
n+1
− n+2
> 0,
1 − 12 αn λn μ 1 − αn+1 λn+1 μ
2 − 2αn λn μ α n λn μ
<1− ,
2 − α n λn μ 2
will be satisfied. So for all n ≥ N = max {n0 , n1 } from (7.18) it follows
 2 2τ λn+1 λ−1
xn+1 − xα  + n+2
xn+1 − yn 2
n+1
1 − αn+1 λn+1 μ
 1 2
αn λ n μ 2 2τ λn λ−1
n+1 2
≤ 1− xn − xαn  + xn − yn−1 
2 1 − α n λn μ

2M (αn+1 − αn )2
+ ,
λn μ αn3
Q.E.D. 
Let us formulate the main result of Chapter.
Theorem 7.1. Let the conditions C1–C7 and D1–D3 are satisfied. Then for
two sequences (xn ), (yn ) generated by iterative Algorithm 7.1 takes place

lim xn − x̄ = lim yn − x̄ = 0, (7.19)

n→∞ n→∞

where x̄ ∈ H is the unique solution of problem (7.1).

Proof. From Lemma 7.3 and inequality (7.13) we have
λn
xn − xαn 2 + 2τ (1 − αn λn μ)−1 xn − yn−1 2 → 0, n → ∞.
λn+1
Hence
lim xn − xαn  = lim yn−1 − xn  = 0. (7.20)
n→∞ n→∞
158 Adaptive Method for the Variational Inequality Problem Over the Set

From inequality
xn − x̄ ≤ xn − xαn  + xαn − x̄ ,
Lemma 7.2 and (7.20) we obtain (7.19). 
Remark 7.5. Easy to see that sequence (zn ) (zn = xn − αn λn F xn ) also
converges strongly to point x̄ ∈ C.

7.6 Algorithm for Two-Level Variational Inequalities

Now we consider a particular case of two-level variational problem (7.1):
two-level variational inequality problem in real Hilbert space H:
find x ∈ V I (F2 , V I (F1 , C)) (7.21)
We assume that the following conditions are satisfied: set C ⊆ H is
closed and convex; operator F1 : C → H is Lipschitz continuous and
monotone; set V I (F1 , C) is non-empty; operator F2 : C → H is strongly
monotone and Lipschitz continuous. Let PC be a metric projection operator
on closed convex set C i.e. PC x is an unique element of C with property
PC x − x = min z − x .
z∈C

For problem (7.21) Algorithm 7.1 takes the following form.

Algorithm 7.2.
) *
Initialization. Choose elements x1 , y0 ∈ C, and numbers τ ∈ 0, 13 , λ1 ∈
(0, +∞). Let n = 1.
Step 1. Calculate
zn = xn − αn λn F2 xn .
Step 2. Calculate
yn = PC (zn − λn F1 yn−1 ) .
Step 3. Calculate
xn+1 = PC (zn − λn F1 yn ) .
Step 4. Calculate
⎧
⎪ λ , if (F1 yn−1 − F1 yn , xn+1 − yn ) ≤ 0,
⎨ n 
λn+1 = τ yn−1 − yn 2 + xn+1 − yn 2
⎪
⎩ min λn , , otherwise.
2 (F1 yn−1 − F1 yn , xn+1 − yn )
 7.7 Conclusion 159

Let n := n + 1 and go to step 1.

From Theorem 7.1 the next result follows.
Theorem 7.2. Let H be a real Hilbert space, C ⊆ H is non-empty closed
convex set; operator F1 : C → H is Lipschitz continuous and monotone; set
V I (F1 , C) is non-empty; operator F2 : C → H is strongly monotone and
Lipschitz continuous; conditions D1–D3 are satisfied. Then two sequences
(xn ) and (yn ) generated by Algorithm 7.2 strongly converge to the unique
solution of problem (7.21).
Remark 7.6. Like theorem 7.2 results takes place for next modification of
Algorithm 7.2 changing the rule of calculation λn on
⎧
⎪
⎨ λn , if F1 yn−1 − F1 yn = 0,
 
λn+1 = yn−1 − yn 
⎪
⎩ min λn , τ , otherwise,
F1 yn−1 − F1 yn 
) *
where τ ∈ 0, 13 .
Remark 7.7. We can build an economic modification of Algorithm 7.2. We
change step 3, letting [19]

xn+1 = PCn (zn − λn F1 yn ) ,

where

Cn = {z ∈ H : (zn − λn F1 yn−1 − yn , z − yn ) ≤ 0} .

7.7 Conclusion
In this chapter we consider a two-level variational problem: a variational
inequality problem over the set of solutions of the equilibrium problem
[22, 23]. To solve this two-level problem, an iterative proximal method is pro-
posed that combines the ideas of a two-stage proximal point method [20, 21],
adaptation and scheme of iterative regularization [1, 27]. The algorithm has
the structure:
⎧
⎪
⎪ zn = xn − αn λn Fxn , 
⎨
yn = arg miny∈C λn G (yn−1 , y) + 12 y − zn 2 ,
⎪  
⎪
⎩ xn+1 = arg miny∈C λn G (yn , y) + 1 y − zn 2 ,
2
160 Adaptive Method for Variational Inequality Problem Over the Set

where we choose λn > 0 adaptively, and positive

αn+1 −αn
sequence (αn ) is such that
lim αn = 0, ∞ n=1 n α = +∞ and lim α2n
= 0.
n→∞ n→∞
In contrast to the rules for choosing the step size used earlier [23], the
proposed proximal algorithm does not calculate the values of the bifunction
at any additional points and does not require knowledge of information about
the Lipschitzian constants of the bifunction.
For monotone bifunctions of Lipschitzian type and strongly monotone
Lipschitz continuous operators, a theorem on the strong convergence of the
algorithm is proved. The proof is based on the results and techniques of [20,
21, 23, 27, 29].
In the future, it is planned to study the possibility of replacing the proxi-
mal steps with constructive instructions such as the steps of the subgradient
method. It is also interesting to relax the rather constraining condition of
monotonicity of the bifunction.
The results obtained are useful in the theory of optimal control. Indeed,
the problem of optimal control of the system can be formulated as a vari-
ational inequality. In the case of non-uniqueness of the optimal control, an
additional criterion can be introduced for the regularization of the problem
and a two-level variational inequality can be obtained. And to solve the
obtained problem, we use the algorithm proposed and studied in this chapter.
In future chapters, we will show in detail how this approach is applied to
problems close to those considered in [10].

Acknowlegdments
This work was supported by Ministry of Education and Science of Ukraine
(project “Mathematical modelling and optimization of dynamical systems for
defense, medicine and ecology”, 0119U100337) and the National Academy
of Sciences of Ukraine (project “New methods of research of correctness and
solution search for discrete optimization problems, variational inequalities
and their applications”, 0119U101608).

References
[1] A.B. Bakushinskii, A.V. Goncharskii, Iterative Methods for Solving Ill-
Posed Problems, Moscow: Nauka, 1989.
[2] F. Browder, “Existence and approximation of solutions of nonlinear
variational inequalities,” Proc. Nat. Acad. Sci. USA, vol. 56, no. 4,
pp. 1080–1086, 1966.
 References 161

[3] F. Browder, “Convergence of approximants of fixed points of nonex-

pansive non-linear mappings in Banach spaces,” Arch. Rational Mech.
Anal., vol. 24, pp. 82–90, 1967.
[4] H. Attouch, “Viscosity Solutions of Minimization Problems,” SIAM J.
Optim., vol. 6, pp. 769–806, 1996.
[5] V.V. Podinovskii, V.M. Gavrilov, Optimization with respect to succes-
sively applied criteria, Moscow: Sovetskoe Radio, 1975.
[6] M. Solodov, “An explicit descent method for bilevel convex optimiza-
tion,” Journal of Convex Analysis, vol. 14, pp. 227–238, 2007.
[7] Kuntsevich, V.M., et al (Eds). Control Systems: Theory and Applica-
tions. Series in Automation, Control and Robotics, River Publishers,
Gistrup, Delft, 2018.
[8] Kondratenko, Y.P., Kuntsevich, V.M., Chikrii, A.A., Gubarev, V.F.
(Eds). Advanced Control Systems: Theory and Applications. Series in
Automation, Control and Robotics, River Publishers, Gistrup, 2021.
[9] Kondratenko, Y.P., Chikrii, A.A., Gubarev, V.F., Kacprzyk, J.
(Eds). Advanced Control Techniques in Complex Engineering Systems:
Theory and Applications. Dedicated to Professor Vsevolod M. Kunt-
sevich. Studies in Systems, Decision and Control, Vol. 203. Cham:
Springer Nature Switzerland AG, 2019.
[10] S.I. Lyashko, D.A. Klyushin, D.A. Nomirovsky, V.V. Semenov, “Iden-
tification of age-structured contamination sources in ground water,”
in: Boucekkine R., Hritonenko N., and Yatsenko Y. (eds.) Optimal
Control of Age-Structured Populations in Economy, Demography,
and the Environment, Routledge, London–New York, pp. 277–292,
2013.
[11] V.V. Kalashnikov, N.I. Kalashnikova, “Solution of two-level variational
inequality,” Cybernetics and Systems Analysis, vol. 30, pp. 623–625,
1994.
[12] I.V. Konnov, “On systems of variational inequalities,” Izvestiya Vuzov,
Matematika, no. 12, pp. 79–88, 1997.
[13] L.D. Popov, “Lexicographic variational inequalities and some appli-
cations,” Math. Programming. Regularization and Approximation. A
Collection of Papers, Tr. IMM, vol. 8, no. 1, pp. 103–115, 2002.
[14] L.D. Popov, “A one-stage method of solving lexicographic variational
inequalities,” Izvestiya Vuzov, Matematika, no. 12, pp. 71–81, 1998.
[15] G. Kassay, V.D. Radulescu, Equilibrium Problems and Applications,
London: Academic Press, 2019.
162 Adaptive Method for Variational Inequality Problem Over the Set

[16] G. Mastroeni, “On auxiliary principle for equilibrium problems,” in

Equilibrium Problems and Variational Models, Daniele, P. et al., Eds.
Dordrecht: Kluwer Academic Publishers, pp. 289–298, 2003.
[17] P.L. Combettes, S.A. Hirstoaga, “Equilibrium Programming in Hilbert
Spaces,” J. Nonlinear Convex Anal., vol. 6, pp. 117–136, 2005.
[18] T.D. Quoc, L.D. Muu, N.V. Hien, “Extragradient algorithms extended
to equilibrium problems,” Optimization, vol. 57, pp. 749–776, 2008.
[19] V.V. Semenov, “Modified Extragradient Method with Bregman Diver-
gence for Variational Inequalities,” Journal of Automation and Informa-
tion Sciences, vol. 50, issue 8, pp. 26–37, 2018.
[20] L. Chabak, V. Semenov, Y. Vedel, “A New Non-Euclidean Proximal
Method for Equilibrium Problems,” in: Chertov O., Mylovanov T.,
Kondratenko Y., Kacprzyk J., Kreinovich V., Stefanuk V. (eds.) Recent
Developments in Data Science and Intelligent Analysis of Informa-
tion. ICDSIAI 2018. Advances in Intelligent Systems and Computing,
vol. 836, Springer, Cham, pp. 50–58, 2019.
[21] D.A. Nomirovskii, B.V. Rublyov, V.V. Semenov, “Convergence of
Two-Stage Method with Bregman Divergence for Solving Vari-
ational Inequalities,” Cybernetics and Systems Analysis, vol. 55,
issue 3, pp. 359–368, 2019.
[22] V.V. Semenov, “Strongly Convergent Algorithms for Variational
Inequality Problem Over the Set of Solutions the Equilibrium Prob-
lems,” in Continuous and Distributed Systems. Solid Mechanics and
Its Applications, vol. 211, M.Z. Zgurovsky, V.A. Sadovnichiy (eds.)
Springer International Publishing Switzerland, pp. 131–146, 2014.
[23] Ya.I. Vedel, S.V. Denisov, V.V. Semenov, “Algorithm for variational
inequality problem over the set of solutions the equilibrium problems,”
Journal of Numerical and Applied Mathematics, no. 1 (133), pp. 5–17,
2020.
[24] G. Gidel, H. Berard, P. Vincent, S. Lacoste-Julien, “A Varia-
tional Inequality Perspective on Generative Adversarial Networks,”
arXiv:1802.10551, 2018.
[25] A. Bohm, M. Sedlmayer, E.R. Csetnek, R.I. Bot, “Two steps at a time –
taking GAN training in stride with Tseng’s method,” arXiv:2006.09033,
2020.
[26] R.I. Bot, M. Sedlmayer, P.T. Vuong, “A relaxed inertial forward-
backward-forward algorithm for solving monotone inclusions with
application to gans,” arXiv:2003.07886, 2020.
 References 163

[27] L.D. Popov, “On schemes for the formation of a master sequence in a
regularized extragradient method for solving variational inequalities,”
Russian Mathematics, vol. 48, iss. 1, pp. 67–76, 2004.
[28] H.H. Bauschke, P.L. Combettes, Convex Analysis and Monotone Oper-
ator Theory in Hilbert Spaces. Berlin, Heidelberg, New York: Springer,
2011.
[29] Y.I. Vedel, S.V. Denisov, V.V. Semenov, “An adaptive algorithm for the
variational inequality over the set of solutions of the equilibrium prob-
lem,” Cybernetics and Systems Analysis, vol. 57, issue 1, pp. 91–100,
2021.
 Part II
Advances in Control Systems Application
 8
Identification of Complex Systems in the
Class of Linear Regression Models

V. Gubarev, S. Melnychuk∗ , and N. Salnikov

Space Research Institute, National Academy of Sciences of Ukraine

and State Space Agency of Ukraine,
40, b. 4/1 Glushkov av., Kyiv, 03187, Ukraine
E-mail: [email protected]; [email protected]
*Corresponding Author:

Abstract
Most methods of analysis and synthesis of control assume the known mathe-
matical model of the control object. To obtain the desired control result, the
model must accurately describe the real processes [1]. It is often not possible
to obtain quantitative relations between the parameters and variables of the
system under consideration on the basis of theoretical laws. For many causal
systems, these relations can only be assessed experimentally, during identifi-
cation. The most developed approach is the so-called stochastic identification,
which consists in obtaining unbiased estimates of the parameters for certain
model. However, the possibilities of such identification are significantly
limited in the case of complex systems. This chapter discusses a more general
approach to identification within the regression model class, which searches
for approximate solutions that are consistent in accuracy with errors in the
data. Since identification problems in complex systems are often incorrectly
posed, the proposed method includes regularization that provides practically
suitable solutions.

Keywords: Identification, complex discrete system, linear regression,

bounded errors, regularization, SVD, regularized dimension.

167
168 Identification of Complex Systems in the Class

8.1 Introduction
System identification, i.e. the problem of constructing its mathematical model
from experimental data has a long history. It is conducted by scientists who
study the behavior of various systems, as well as control specialists who
need mathematical models to solve problems of analysis and synthesis. Many
results on this problem within the framework of a discrete description have
been published in well-known books [2, 3], including monographs [4, 5].
The most interesting problems from a practical point of view are associ-
ated with uncertainty in the output and input variables. The corresponding
type of identification problem is called “errors in variables” (EIV) and is
considered by many scientists [6–9]. Recent publications can be found in
[10–13]. The main focus there was on obtaining unbiased parameter estimates
and the issue of consistency.
Different approaches were considered within EIV identification, for
example, for describing an uncertain dynamic system it was proposed to use a
graph metric introduced in [14] with the representation of model uncertainties
by the graph subspace consisting of the input and output spaces of the system.
Moreover, it was assumed everywhere that the errors are independent and
identically distributed (i.i.d.) measurement noises. Recently graph subspace
approach was proposed and developed in [15] for EIV identification problem
and a new estimation algorithm was derived that is more general than total
last square algorithm used in preceding works. It is aimed at estimation of the
model parameters based on finite measurement samples of the graph subspace
of a given system.
This chapter explores a different approach to address the EIV identifica-
tion problem with a different interpretation of the uncertainty in the data. We
assume that the errors are randomly distributed but limited in magnitude. As
in [15], the data series can be large, but always finite. This initial information
about the system is more realistic than in the stochastic case.
The approach proposed here is focused on the identification of controlled
systems and allows the use of a fairly arbitrary input signal. This allows data
collection in active experiments with specially designed informative input
signals. The problems of identifying systems of large dimension, including
infinite ones, based on inaccurate data, are usually ill-posed. Obtaining unbi-
ased estimates of the model parameters in these cases becomes impossible.
Moreover, a regularized solution in ill-posed case can give an approximate
model of a lower dimension than that of the original system. Thus, the
proposed identification method involves determining the size of the model
that guarantees stability, followed by parameter estimation.
 8.2 Active Experiments and Informative Data 169

The resulting models were tested when simulating the identification pro-
cess. This chapter is actually developing an approach and results obtained
earlier in [16, 17].

8.2 Active Experiments and Informative Data

This chapter focuses on systems of large or infinite dimension, for which
obtaining unbiased estimates is meaningless. We assume that the identified
model may have a smaller dimension than the original system.
The use of a reduced dimension is most often dictated by the very
formulation of the EIV identification. The presence of errors in the initial data
of the identification problem, in many cases, turns it into an ill posed one. As
a result, the identification of any method becomes very sensitive to random
errors in the input data. Re-identification with other errors leads to signifi-
cantly different parameter estimates, so such solutions are practically useless.
Therefore, in identification, it is proposed to use a regularization that
ensures the stability of the obtained solutions. We find and use only those
models whose parameters are less sensitive to errors. For this it is enough to
find approximate solutions that are consistent in accuracy with the errors in
the data.
Naturally, in this case, the accuracy of the resulting models will be
limited. Under such inevitable circumstances, or rather, the fundamental
properties of identification, it is proposed to take the dimension of the sought
model as a regularization parameter. The largest dimension of the model,
admissible according to the stability condition, will correspond to the regu-
larized approximate solution. It should be noted that despite the discreteness
of the regularization parameter, there is no clear boundary between stable and
unstable solutions.
In principle, it is possible to complicate the problem by finding the
optimal or quasi-optimal dimension, taking into account the corresponding
quality criteria of the model. However, this will greatly complicate the
method of its solution, which is undesirable for engineering practice.
Here we consider fairly simple methods for choosing the dimension of
the model, which are analogs of the residual principle and quasi-optimal
regularization used in methods for solving ill-posed problems. The rejection
of unbiased estimates of the stochastic approach allows instead to consider
other non-stochastic formulations of the problem and the corresponding solu-
tion methods, which, nevertheless, retain the basic techniques of stochastic
identification.
170 Identification of Complex Systems in the Class

Instead of interpreting the error in the data as i.i.d, we consider them

arbitrary, but belonging to sets limited to a known value. We restrict consid-
eration to the scalar case, when the object under study has single input u(t)
and single output y(t) (t–discrete moments of time). In cases of identification
of multidimensional systems, various approaches based on the scalar case can
be used. They are not considered in this chapter.
So, instead of exact values of signals u(t) and y(t) we have their
measurements

ũ(t) = u(t) + ξu (t), ỹ(t) = y(t) + ξy (t), (8.1)

where ξu (t) and ξy (t) are errors in data. According to the non-stochastic
approach, we assume that ξu (t) and ξy (t) are unknown arbitrary random
sequences satisfying condition

|ξu (t)| ≤ εu , |ξy (t)| ≤ εy (8.2)

where εu and εy are sufficiently small compared to the magnitude of u(t) and
y(t).
The kind of input signal affects the information contained in the observa-
tions. The most important characteristic of an input is its ability to excite all
modes of the system. Insufficiently excited modes cannot be identified.
For stochastic identification, the persistently exiting input sequence of a
given order [18] is used. An ergodic sequence u(t) is persistently exiting of
order l if and only if
 
1
rank lim UlN · UlN = l,
T
(8.3)
N →∞ N

where
⎡ ⎤
u(t) u(t + 1) . . . u(t + N − 1)
⎢ u(t + 1) u(t + 2) . . . u(t + N ) ⎥
⎢ ⎥
UlN = ⎢ .. .. .. .. ⎥.
⎣ . . . . ⎦
u(t + l − 1) u(t + l) . . . u(t + N + l − 2)

When u(t) is a zero-mean discrete white noise sequence it has the statistical
property
1
UlN · UlN = σu2 · Il + ε2N · EN , (8.4)
N
 8.2 Active Experiments and Informative Data 171

where “  “ is matrix transpose, σu –dispersion, Il is l–dimensional unit matrix,

εN is a sequence such that lim εN = 0 and EN is a matrix of appropriate
N →∞
dimensions with EN  ≤ 1.
In the case of non-stochastic identification with bounded uncertainty
(8.1), (8.2), we plan an active experiment so that each mode of the system
at a certain moment of observation makes the most significant contribution to
the output signal, as much as possible.
The chapter examines the identification of only asymptotically stable
systems. An initial informative data for identification can be obtained in
two ways. The first one is to collect data from many separate experiments,
consisting of the finite intervals of excitation by u(t) = 0 and relaxation with
u(t) = 0 of the same length l, where l should to be larger then duration of
the transient process. An alternative option is a single continuous experiment,
which is compound from previous mentioned, where l should guarantee the
process is completely damped or the steady-state regime is reached.
To select an informative input excitation, consider a discrete linear sta-
tionary system. Such systems can be represented by models in the classes of
linear regression or systems of difference equations in the state space, which
are known to be equivalent.
If we take a state-space model

x(k + 1) = Ax(k) + bu(k)

y(k) = c x(k) + du(k)

with a Jordan block realization A = diag(Ap ), b = col (bp ) , c = col (cp ) ,

where
   c   c 
αp −βp bp cp
Ap = , bp = s , cp = s , λp = αp ± iβp ,
βp αp ) c* pb ) c* c p
Ap = (αp ) , b p = bp , c p = c p , λp = αp ,
(8.5)
and λp are eigenvalues of matrix A, then for controlled and observable
systems we can write the input-output ratio in which the parameters are
invariants. In the case y(t) = 0 and u(t) = 0 at t ≤ 0 and u(0) = 0 at
t > 0, according to [19], this relation is

!
t−1 !
P
y(t) = h0p (t − j)u(j), (8.6)
j=0 p=1
172 Identification of Complex Systems in the Class

where h0p (k) = ρkp [fpc cos ωp (k) + fps sin ωp (k)], and

ρp = |λp | , ωp = arg(λp ), fpc = ccp bcp + csp bsp , fps = ccp bsp − csp bcp .

Parameters ρp , ωp , fpc and fps are invariants characterizing the dynamic

properties of the system. It was assumed in the expression that there are
no multiple roots, and the case of real roots is a special case of (8.5)
corresponding to ωp ≡ 0, βp = 0, csp = bsp = 0.
When identifying an EIV, the informativeness of the input action can be
judged by the signal-to-noise ratio (SNR). The higher SNR for the output
of each mode p (8.5), the more informative is the input action. Therefore,
we will shape input at the excitation intervals so that at its end, one of the
modes has the largest output signal. Different u(t) across experiments should
effectively excite all modes of the system.
We will use 2 types of input signals. The first type is a rectangular pulse

u(t) = 1, t = [1, τ ] , (8.7)

where τ is duration, variable across experiments in the range: τ =

[1, ..., τmax ]. Short ones provide the greatest excitation of fast real modes.
Increasing the duration will add significant signals from slower modes,
regardless of the value of ccp bcp . Therefore, rectangular pulses of different
duration should form a set of initial conditions with different contributions
of separate modes.
The second type of input is a single harmonic

u(t) = sin (ωt) , t = [1, τ ] , (8.8)

where frequency ω changes across experiments, limited by discreteness of

measurements to the range (0, π/2]. For complex modes in (8.5), we aim
to use the resonant excitation when ω happens to be close to some natu-
ral frequency ωp . When ω are varied with a small step, it is possible to
provide resonant excitation of all oscillating modes of the system. In this
case, the duration τ is chosen to be sufficient to establish steady-state forced
oscillations.
Let us denote the step of the rectangular pulse duration by Δ0 , and the
step of the harmonic signal frequency ω by Δ1 . With an appropriate choice
of parameters Δ0 , τmax , Δ1 , and l, it is possible to obtain informative data
with an acceptable SNR for identification with a fairly general understanding
of the system under study.
 8.3 Method of Identification 173

8.3 Method of Identification

In this chapter, we will identify a system in the class of linear autoregressive
models. When considering complex systems that in this class have a very
large or infinite dimension, we come to the need for structural-parametric
identification.
When identifying EIV, an approximate model must be found that is
consistent with data errors and computational accuracy. Then, in the general
case, the dimension of the model may be less than the order of the true system.
Therefore, the identification method should include a procedure for finding an
acceptable dimension of the model with subsequent parametric identification.
In the class of autoregressive models

y(t) = −a1 y(t − 1) − a2 y(t − 2) − . . . − an y(t − n)+

(8.9)
+b0 u(t) + b1 u(t − 1) + . . . + bm u(t − m)

the above-described design of active experiment with excitation and relax-

ation allows splitting the identification problem. Primarily, we estimate
dimension n and autoregressive coefficients a = (−a1 , −a2 , . . . , −an )
from measured output during relaxation, where u(t) ≡ 0. By subtracting
the calculated free motion from the measured output, we obtain the forced
motion data. Finally we estimate dimension m and control coefficients b =
(b0 , b1 , . . . , bm ) from input and corresponding forced motion.

8.3.1 Model Order Selection

At the first stage of the method, the dimension of the model n is determined.
For this, from the measured output data on the relaxation intervals, we form
the following matrix
⎡ pulse pulse ⎤
ỹ1 (τ1 + 1) . . . ỹ1pulse (τ1pulse + l)
⎢ ⎥
⎢ .. .. .. ⎥
⎢ . . . ⎥
⎢ ⎥
⎢ pulse pulse ⎥
⎢ ỹk (τk + 1) . . . ỹ pulse pulse
(τ + l) ⎥
Ỹrelax = ⎢ 1
⎢ harm harm
1 k 1 k 1 ⎥,
⎥ (8.10)
⎢ ỹ1 (τ1 + 1) . . . ỹ1harm (τ1harm + l) ⎥
⎢ ⎥
⎢ .. .. .. ⎥
⎢ . . . ⎥
⎣ ⎦
ỹkharm
2
(τkharm
2
+ 1) . . . ỹkharm
2
(τkharm
2
+ l)
174 Identification of Complex Systems in the Class
−−→
where ỹipulse (t), i = 1, k1 –output from ith rectangular pulse, τipulse –
duration of ith rectangular pulse, k1 - total number of rectangular pulses,
−−→
ỹjharm (t), j = 1, k2 –output from jth harmonic input, τjharm –duration of
jth harmonic input, k2 –total number of input harmonics. We denote the total
number of rows by R = k1 + k2 .
By its structure, the matrix Ỹrelax is not Hankel one, but in a certain sense
it can be considered as its generalization. We use the SVD decomposition

Ỹrelax = U ΣV T , (8.11)

where U , Σ, V are matrices of dimensions R ×R, R ×l and l ×l respectively.

For a system with a finite dimension n, noiseless measurements y(t), R <
n and l < n , the matrix Yrelax will be of incomplete rank, and only the
first n singular numbers σ1 , . . . , σn will be non-zero. Thus, the dimension of
the system can be precisely determined by the number of nonzero diagonal
elements of Σ.
Since measurements ỹ(t) contain uncertainty, the matrix Ỹrelax with
probability almost one will be of full rank and all singular numbers will be
positive. However, with small errors bound εy , it makes sense to divide it
into “signal” and “noise” part. To obtain a model of a given dimension n̂, we
use the low-rank approximation [20] of Ỹrelax . We split matrices (8.11) into
blocks  
8 9 ΣS 0 8 S N 9
U ΣV = U S U N N V V , (8.12)
0 Σ
so matrix ΣS includes first n̂ singular values. Here indices “S” and “N” denote
“signal” and “noise” parts. As a result, we get decomposition
S N
Ỹrelax = Yrelax (n̂) + Yrelax (n̂) , (8.13)
S
where Yrelax (n̂) = U S ΣS V S and Yrelax
N (n̂) = U N ΣN V N . Expansion
(8.13) allows, proceeding from the ε–rank property [20] to find the matrix
of the given rank closest to full-rank ones in Frobenius norm:
) S *  N 
rank Yrelax (n̂) = n̂, Yrelax (n̂)F = σn̂+1 , (8.14)

where  · F means Frobenius norm. The value σn̂+1 can be used to judge
the proximity of the matrix Ỹrelax to the matrix of rank n̂. Sometimes, the
true dimension n can be determined by the rate of decrease of the singular
values. This can be done if there is a pronounced gap between the values
of σi and σi+1 . However, there is often no clear gap. It all depends on
 8.3 Method of Identification 175

the signal-to-noise ratio of individual modes (8.5). Modes with a low SNR
becomes indistinguishable from background noise, making impossible in the
non-stochastic case to establish the exact dimension of the system. In such
cases, we can find a model of reduced dimension. For this, the principle of
regularization is used. The required approximate solution must be consistent
in accuracy with the uncertainty of the data.
Consider how the quantities σn̂+1 and εy are related. Let’s represent the
matrix Ỹrelax in the form of the exact and noise part, omitting here the “relax”
index
Ỹ = Y + ΔY. (8.15)
Inequality (8.2) gives ΔY max < εY . Let us estimate from above the
   
quantity σn̂+1 . From (8.14) we have σn̂+1 = Y N (n̂) = Ỹ − Y S (n̂) .
F F
Substituting (8.15) and using the triangle inequality we derive
 
σn̂+1 ≤ Y − Y S (n̂)F + ΔY F .
   
 
Since the quantity Y − Y S (n̂)F can not be larger then Y − Ỹ  , we
F
have
σn̂+1 ≤ 2 ΔY F .

 
N M 2
From norm equivalence AF = j ai,j , Amax =
& & i
& &
max &ai,j &, we have estimation
i = 1, M
j = 1, N
√
σn̂+1 ≤ 2 Rl · εY . (8.16)
Relation (8.16) means that the modes of the system, starting with such n̂,
cannot be guaranteed to be identified. The possibility of their identification is
determined only by the case of noise realization. This property will be used to
determine the dimension of the model, namely the dimension of the vector a.
In certain cases, by reducing Δ0 and Δ1 , and, consequently, the magni-
tude of R, it is possible to achieve an approximation to the exact dimension.
Everything will depend on the dynamic properties of the system, determined
by its expansion invariants (8.5). Low SNR values appear for modes with
close ωp or modes with small fpc and fps . There exist those εy , starting from
which their signals are not distinguishable against the background of noise.
The presence of such modes in the system will lead to the reduction of the
model order in the proposed approach.
176 Identification of Complex Systems in the Class

8.3.2 SVD Parametric Identification

After establishing the dimension n of the autoregressive part of the model, the
problem of determining the parameters a is being solved. The data of matrix
Y S (n) is used to form an overdetermined system of linear equations.
Y A (n)a = ȳ A (n). (8.17)
The matrix Y A (n) in (8.17) is formed from all equations of type
(y r (t − 1), y r (t − 2), . . . , y r (t − n)) · a = −y r (t), (8.18)
−−→
where index r denotes row from matrix Y S (n), r = 1, R and t goes through
all possible values from n+1 up to l. When forming the system, the equations
are filtered. If the absolute values of y are less than a certain threshold level
K N S · εY ,
then such an equation is discarded. This is done in order to ignore noisy data.
The vector a is found from (8.17) by the usual or generalized least squares
method (LS) [20]. However, it should first be sure that the matrix is not ill-
conditioned. An estimate of the condition number in the norm || · ||2 can be
established from the ratio of the singular values of the matrix Y A
σ1
æ2 (n) = . (8.19)
σn
If the condition number (8.14) is acceptable, then the solution is found by
standard, for example, generalized LS. Otherwise, it is necessary to use the
regularized LS, which is available in the libraries of standard programs.

8.3.3 Total Model Reconstruction

The described procedure for finding the dimension and coefficients of a
vector a is fully applicable for the vector b. As a result, we will have a
complete solution to the identification problem. It should be noted here that
the dimension of the vector b cannot exceed n + 1, when the observation
equation has the form y(t) = c x(t) + du(t) and d = 0, and the value n
when d = 0. In the latter case, the dimension of b may be smaller than n. The
task of finding b is preceded by a procedure for extracting a purely forced
motion from the sequence {y(t)}. The estimated output signal of purely
˜ is determined by the relation
forced motion ỹ(t)
˜ = ỹ(t) + a1 ỹ(t − 1) + a2 ỹ(t − 2) + . . . + an ỹ(t − n).
ỹ(t)
 8.3 Method of Identification 177

Then the basic equation for finding the vector b will be the following equation
˜ = b0 u(t) + b1 u(t − 1) + . . . + bm u(t), d = 0
ỹ(t)
˜ = b1 u(t − 1) + . . . + bm u(t), , (8.20)
ỹ(t) d=0
for any moment of time t where we have nonzero excitation signals. From
the data on the intervals of excitation of a suitable duration, we form an
overdetermined system
˜
W · b = ỹ, (8.21)
where b is a vector whose components are b0 , b1 , . . . , bm or
b1 , b2 , . . . , bm , the vector ỹ˜ is composed of ỹ(t)
˜ for the time moments
corresponding to (8.20), and W is a matrix consisting of the input signals for
the appropriate time moments. In (8.21), both the matrix W and vector ỹ(t) ˜
are given approximately. Taking into account the specified errors, we will
find an approximate solution of (8.21), using the generalized LS. In this case,
you should make sure that the problem being solved is correctly posed. If it
is ill-conditioned, a regularized solution should be found in accordance with
the well-known methods for solving incorrect linear systems. In this case, it
is possible to obtain stable approximate solutions, like with a vector a, by
lowering the dimension m of the vector b.
Bad conditionality of the matrix W indicates that the dimension of the
vector b is less than n. In such cases, the further actions are similar to those
described in Section 2.1, when determining the dimension of the vector a. To
do this, we carry out the SVD decomposition of the matrix W , and check for
singular values. After establishing the dimension of the vector b its value is
found from the redefined system (8.21).
Solving the system of equations (8.21) by generalized LS completes the
system identification, resulting in an approximate model consistent with the
EIV.

8.3.4 Quasioptimal Model Dimensions

The dimensions of vectors a and b defined by cited above can be considered
as using the residual principle for this purpose [21]. In addition to it, it is
permissible to find the dimension of the model based on a method that allows
calculating the quasi-optimal value of the regularization parameter, in our
case the dimension of the model [21].
The criterion (8.16) for determining the dimension does not guarantee
that the coefficients of the found model will be determined consistently, even
if the exact dimension is determined.
178 Identification of Complex Systems in the Class

Then the quasi-optimal dimension should be found. For the varied n in

decreasing order the problem of parametric identification is solved and the
variation of the coefficients are estimated
   
1   
ˆ(n)]  ,  1 [b(n + 1) − ˆb(n)]  ,
 [a(n + 1) − a (8.22)
n  n 

where a(n + 1) and b(n + 1) is the solution to the problem of parametric

ˆ(n) and ˆb(n)
identification of vectors a and b for dimension n + 1, and a
are extended vectors formed from solutions for dimension n according to the
following rule
ˆ(n) = (−a1 − a2 . . . − an 0) = (a (n) 0) ,
a

ˆb(n + 1) = (b1 b2 . . . bn 0) = (b (n) 0) ,

where a(n) and b(n) are the values of the corresponding vectors for the
dimension n. Various behaviors of the discrete function defined by (8.22)
are possible. At first, it decreases, and then begins to grow, while local
minima are possible. Then the last local minimum determines the quasi-
optimal dimension. It is possible when this function grows continuously,
starting from the smallest n. It is also possible a noticeable scatter in the
behavior of this function, starting from some point. Then the value n before
the scatter determines the desired dimension.
The parameters a and b corresponding to the quasi-optimal n determine a
regularized approximate solution to the identification problem.

8.4 Simulation Results

The purpose of the model experiments was to demonstrate the efficiency of
the proposed approach to identification and the method described, as well
as to evaluate its effectiveness by examining the quality of the resulting
models and the features that arise when solving specific problems. Since
the result of identification depends on the invariants of the systems, the
capabilities of the method have been demonstrated on the most suitable
systems for identification. This allows one to assess the limiting capabilities
of the method.
The most suitable form for defining a system with specified dynamic
properties is a representation in the form (8.5). Using it we can choose invari-
ants that provide the best conditions for identification. So, we set eigenvalues
 8.4 Simulation Results 179

λp for all modes, ensuring stability, slow decay and sufficient distance from
each other. We also set invariants of |fpc | and |fps | to be equal for all modes, to
provide better observability. The definition of the system in the form (8.5) in
canonical observable representation (ccp = 1, csp = 0) makes it easy to pass
to equivalent description in a normal observable realization in the state space
⎡ ⎤
0 0 . . . 0 −an ⎡ ⎤ ⎡ ⎤
⎢ 1 0 . . . 0 −an−1 ⎥ bn 0
⎢ ⎥ ⎢ bn−1 ⎥ ⎢ 0 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
A = ⎢ 0 1 . . . 0 −an−2 ⎥ , b = ⎢ . ⎥ , c = ⎢ .. ⎥ .
⎢ .. .. . . .. .. ⎥ ⎣ .. ⎦ ⎣ . ⎦
⎣ . . . . . ⎦
b1 1
0 0 . . . 1 −a1

and then to an equivalent representation in the form of autoregression (8.9):

ai = ai , bi = dai + bi , i = 1, n, b0 = d, if m=n.
Thus, we have the opportunity to investigate the identification method for
systems with different dynamic properties. It is also important to establish the
impact on the quality of models of errors in the data within the constraints
defined by (8.1), (8.2). We assume that the values of εu and εy are also quite
small compared to deterministic signals. To correctly compare and evaluate
simulation results, we will consider the relative levels of errors
ξu max ξy max
= = ε,
umax ymax
where ε is a variable value. Taking into account constraints (8.2), it leads to
the relations
εu = u max · ε, εy = ymax · ε. (8.23)
Then εu and εy characterize the relative width of the intervals of membership
of errors, and ||u||∞ and ||y||∞ characterize the same for input and output
signals. Using (8.22), the dependence of the quality of the resulting models
on ε was investigated.

8.4.1 Algorithm of identification

1. A series of active experiments with the excitation and relaxation:
8 9
upulse
i (t) = 1, t = 1, τi
8 9 −−→
upulse
i (t) = 0, t = τi + 1, τi + l τi = i, i = 1, k1 ,
180 Identification of Complex Systems in the Class
8 9
uharm
i (t) = sin (ωi t) , t = 1, τi ωi = 0.01i
8 9
uharm
i (t) = 0, t = τi + 1, τi + l τi = 10 · min (10, π/ωi ),
−−→
i = 1, k2 ,

where k1 = 100, k2 = 157, l = 1000. The noisy outputs are stored line
by line in the matrix Ỹ , the inputs are stored line by line in the matrix Ũ .
2. From Ỹ on the relaxation, a matrix Ỹrelax ∈ RR×l is formed. R =
k1 + k 2 .
3. Using the SVD, dimension n is determined.
4. S
Matrix Ỹrelax is splitted by n into signal Yrelax N
(n) and noise Yrelax (n)
parts.
5. A linear system for the regression coefficients a is formed from
S
Yrelax (n). Equation
  included if its coefficients exceed the SNR thresh-
 
old KN S · ε · Ỹ  .
max
6. The regression coefficients of free motion are found.
7. We define forced movement Ỹf orced by subtracting free movement from
Ỹ . ' (
8. (optional). We build compound matrix Ỹf orced |Ũ and apply SVD.
If it is not possible to determine dimension m by 'singular values, (
then we accept m = n. Extract signal part of Ỹf orced |Ũ =
' (S ' (N
Ỹf orced |Ũ (m̂) + Ỹf orced |Ũ (m̂), and then break it back:
' (S ' (
Ỹf orced |Ũ = ỸfSorced |Ũ S .
9. From the matrices Ỹf orced and Ũ (or ỸfSorced and Ũ S ) we form a system
for the control coefficients b of regression. The system includes all
equations whose coefficients exceed the threshold value, similarly to
item 5.
10. By singular values we check whether the chosen dimension m̂ is true. If
not, we rebuild the system for the specified dimension.
11. We solve the system and find the regression coefficients for control.
Then we check regression model:
1. Check if the resulting model is stable.
2. Calculate the deviation of the model output from the real system when
testing with a rectangular pulse followed by relaxation.
 8.4 Simulation Results 181

8.4.2 Excitation of system modes

Consider the system as a set of individual modes (8.5), each of which corre-
sponds to either a real eigenvalue or a pair of complex conjugate eigenvalues.
The mode can be identified only when its contribution to the output is
distinguishable against the noise. When u = 0 it is defined as yp = xp Acp
for real λp , and yp = xcp Acp + xsp Asp for complex λp , where p is a mode
) *T
number, xp and xp = xcp , xsp are states of real and complex modes. & &
For real λp , the magnitude of the mode output is proportional to the &xp &.
&Therefore,
& the best conditions for identifiability correspond to maximizing
&xp &. For the complex mode, the magnitude of yp is determined by the dot
) *
product of xp and the vector of coefficients A  p = Ac , As T . The best
p p
conditions for identifiability correspond to a minimizing the angle between
xp and A  p and maximizing xp  . Vector A  p is unknown in advance. Free
2
movement of a separate complex mode for λ = α ± iβ is described by the
equation  c    c 
xk+1 α −β xk
= ,
xsk+1 β α xsk
where the subscript indicates the quantization step number. We represent the
eigenvalue
< in form α + iβ = reiϕ , where α = r cos ϕ, β = r sin ϕ, r =
α2 + β 2 < 1. Then the transformation matrix of the state vector
   
α −β cos ϕ − sin ϕ
=r
β α sin ϕ cos ϕ

is a combination of rotation by an angle ϕ and scaling with a factor r. The end

of the vector xp in free motion describes a contracting spiral on the Oxc xs
plane, and the angle between xp and A  p at each turn approaches to 0 and π
ϕ
closer than /2. The maximum mode output will be reached at the first of
these points and is estimated as
'
π/ (+1 ) *
x0 2 · r ϕ · c2 · cos ϕ/2 ≤ ymax ,

where x0 is the state vector at the first relaxation step. Therefore, the best
conditions for identifying the mode correspond to maximization x2 for any
implementation of A.
Conclusion: the magnitude of the mode excitation will be defined as the norm
of the state vector.
182 Identification of Complex Systems in the Class

Figure 8.1 State of a complex mode under harmonic excitation with different initial phases

Let us consider the behavior of the state vector of a complex mode upon
excitation by a harmonic signal with a different initial phase. Figure 8.1 shows

trajectory (xc , xs ) of mode 0.95e0.5i for excitations u (t) = sin (0.2t) (left),
and u (t) = cos (0.2t) (right). In both cases, the limiting trajectories coincide.
Conclusion: the steady-state forced motion of the complex mode does not
depend on the initial phase of the harmonic signal.
Consider the excitation of a complex mode by harmonic signals of differ-

ent frequencies. Figure 8.2 shows trajectory (xc , xs ) of mode 0.95e0.5i for
excitations u (t) = sin (0.5t) (left), and u (t) = cos (0.8t) (right). The first
case is resonant.
The steady-state forced motion of the complex mode under harmonic
excitation has trajectory of the form of an ellipse and depends on the ratio
of the natural frequency and the input frequency. In the resonant case, the
trajectory is close to circular.
Let us consider the trajectories of free motion after the excitation of
a mode by a harmonic signal. We use harmonic excitation and turn off
control at times corresponding to different points of the elliptical trajectory.

Figure 8.3 shows trajectory (xc , xs ) of mode 0.95e0.5i for input u (t) =
 −→
sin (0.2t) , t = 1, τ
for τ = 196 (solid) and τ = 204 (dashed).
0, t>τ
Conclusion: in the case of nonresonant harmonic excitation of the complex
mode, the value of the relaxation signal can significantly depend on the initial
moment.
 8.4 Simulation Results 183

Figure 8.2 State of a complex mode under harmonic excitation of different frequencies

Figure 8.3 State of a complex mode during relaxation after harmonic excitation

Let us now consider the excitation of a complex mode by a constant


signal. Figure 8.4 shows trajectory (xc , xs ) of mode 0.95e0.5i for u (t) =
1, t ≤ τ (solid) and u (t) = 0, t > τ (dashed).
Conclusion: constant excitation moves the state of the complex mode along a
contracting spiral to a certain limiting value. When the control is turned off,
it spirals back to zero.
184 Identification of Complex Systems in the Class

Figure 8.4 State of a complex mode during constant excitation and relaxation

Figure 8.5 The norm of the state vector of modes under harmonic excitation

Let us consider the change in the norm of the state vector of various
modes.
6 We will take a system of7 two real and two complex modes: {λ} =
0.5, 0.9, 0.95e0.5i , 0.95e1.2i . Figure 8.5 shows dependency xp  and
|xp | on time for input u (t) = sin (0.2t).
Excitation of the system by a harmonic input signal leads all modes to
forced oscillations with the same frequency, but out of phase.
Figure 8.6 shows dependency xp  and |xp | on time for input u (t) ≡ 1.
 8.4 Simulation Results 185

Figure 8.6 The norm of the state vector of modes under constant excitation

Excitation of the system by a constant input signal leads the modes to a

certain limiting excited state.
Figure 8.7 shows dependency xp  and |xp | on time for rectangular pulse
with length 20.
When the control is turned off, the norm of the mode state vector
decreases exponentially.

Figure 8.7 The norm of the state vector of modes under rectangular pulse and on relaxation
186 Identification of Complex Systems in the Class

Figure 8.8 The maximum of norm of the modes state vectors under harmonical input

Let us consider to what extent the modes can be excited as much as

possible under harmonic excitation. We will apply harmonics to the input
of the same system by changing the frequency. Figure)8.8 shows
* dependency
maximum of xp  and |xp | on input frequency ωU ∈ 0, π/2 .
Complex modes are better excited on resonant frequencies. Real modes
are better excited with low frequencies.
Simulation has shown that the described set of input signals of harmonic
and rectangular pulses allows maximum excitation of all modes in separate
experiments.

8.4.3 Effect of Using the Signal Part of Matrices

Two alternatives to the identification algorithm are considered:
S
1. To use or not to use the signal matrix Yrelax (n) instead of matrix Ỹrelax
when forming the system of the regression coefficients a.
2. To use or not to use the signal matrices ỸfSorced and Ũ S instead of
matrices Ỹf orced and Ũ when forming the system of the regression
coefficients b.
We will conduct identification of a system, but apply the following configu-
rations of the algorithm:
A. Do not use signal matrix at all.
 8.4 Simulation Results 187

B. Use when finding a.

C. Use when finding a and b.
6 7
We use system with modes {λ} = 0.5, 0.9, 0.95e0.5i , threshold for
equations KN S = 100 and 100 random realizations of noise with relative
error εU = 10−4 , εY = 10−4 and εU = 10−6 , εY = 10−4 . The quality of
the model was checked by the error in estimated regression coefficients and
by the model’s response to a rectangular pulse of duration 50. Results are
represented in Tables 8.1 and 8.2.

Table 8.1 Maximum and average errors of models for εU = 10−4 and εY = 10−4
Algorithm configuration Error in a Error in b Error by output
A 0.02263 (0.01980) 0.19602 (0.16965) 0.01854 (0.01625)
B 0.00281 (0.00091) 0.02635 (0.00853) 0.00234 (0.00074)
C 0.00281 (0.00091) 0.03752 (0.02049) 0.00332 (0.00225)

Table 8.2 Maximum and average errors of models for εU = 10−6 and εY = 10−4
Algorithm configuration Error in a Error in b Error by output
A 0.02246 (0.01985) 0.19539 (0.17013) 0.01835 (0.01631)
B 0.00283 (0.00092) 0.02718 (0.00820) 0.00219 (0.00074)
C 0.00283 (0.00092) 0.02239 (0.00677) 0.00216 (0.00068)

S
The extraction of a matrix Yrelax (n) of a given rank from the matrix
Ỹrelax can significantly improve the quality of the resulting model. The
selection of a matrices ỸfSorced and Ũ S from matrices Ỹf orced and Ũ has little
effect on the quality of the resulting model. When εU is relatively small,
we can achieve some improvement. In other case we risk worsening the
result.

8.4.4 Effect of Discarding Data Close to Noise

Here we consider the effect of discarding rows from a system of equations
that are formed from noisy data with a low signal-to-noise ratio. This happens
when we form a system from YrelaxS (n) to find a, and when we form a system

from Ỹf orced and Ũ to find b. 6 7
We use system with modes {λ} = 0.5, 0.9, 0.95e0.5i , variable
−−−→
threshold KN S = 5, 500 and random noise with relative error εU = 10−4 ,
εY = 10−4 . The quality of the model was checked by the model’s response
to a rectangular pulse of duration 50.
188 Identification of Complex Systems in the Class

Figure 8.9 Average number of rows in overdetermined systems for a and b (left), the quality
of the resulting model for different implementations of random noise (right)

Figure 8.9 shows the number of not discarded equations (left) and model
quality (right) for varied coefficient KN S .
The right figure shows the dependencies for four different variants of
the implementation of random measurement noise. In one of these cases,
dropping a large number of equations did not improve the result. An increase
in the threshold SNR coefficient can significantly reduce the number of lines
in the systems of equations and, in most cases, slightly increase the accuracy
of the model.

8.4.5 The Best Systems for Identification

Let us carry out identification modeling for different systems of the same
order to determine those of them that are identified most accurately.
We use threshold KN S = 100 and noise with relative error εU = 10−5 ,
εY = 10−5 .
Results are represented in Table 8.3.
Complex modes are identified better than real ones. The presence of close
real or complex eigenvalues degrades the identification accuracy. The best
system for identification is purely oscillatory, weakly damped, and does not
contain close natural frequencies.

8.4.6 Model Order Determination

We carry out an experiment to determine the dimension of the system in terms
of singular numbers. We use favorable configurations as initial systems: all
 8.4 Simulation Results 189

Table 8.3 The worst and average quality of models

Modes of a system Error in a Error in b Error by output

0.95e0.5i , 0.95e1.2i 2x10−6 (1x10−6 ) 5.6x10−5 (2.1x10−5 ) 5x10−6 (3x10−6 )

0.95e0.1i , 0.95e1.2i 4x10−6 (1x10−6 ) 4.3x10−5 (1.5x10−5 ) 1.9x10−5 (7x10−6 )
0.95e0.5i , 0.95e0.6i 3x10−5 (8x10−6 ) 5.7x10−5 (1.8x10−5 ) 2.6x10−4 (1.1x10−4 )
0.8e0.5i , 0.8e0.6i 2.7x10−4 (1x10−4 ) 5.2x10−4 (1.7x10−4 ) 2.2x10−3 (6x10−4 )

0.2, 0.9, 0.95e0.5i 0.00046 (0.00010) 0.00054 (0.00012) 0.00007 (0.00002)

0.5, 0.9, 0.95e0.5i 0.00031 (0.00009) 0.00269 (0.00086) 0.00024 (0.00007)
0.2, 0.5, 0.95e0.1i 0.00217 (0.00070) 0.00165 (0.00055) 0.00069 (0.00024)
0.2, 0.3, 0.95e0.5i 0.02219 (0.00825) 0.02133 (0.00794) 0.00676 (0.00253)

0.2, 0.5, 0.85, 0.9 0.00682 (0.00227) 0.00629 (0.00208) 0.00363 (0.00117)
0.2, 0.5, 0.70, 0.9 0.01780 (0.00495) 0.01673 (0.00474) 0.00425 (0.00132)
0.2, 0.5, 0.89, 0.9 0.07777 (0.05265) 0.07756 (0.05248) 0.13216 (0.09800)

modes are complex, and are defined as

 
p−1 −−→
λp = 0.95eiϕp , ϕp = 0.3 + 1.2 · , p = 1, P ,
P −1
that means the uniform distribution of p natural frequencies in the interval
[0.3, 1.5]. The dimension n of the autoregressive part is equal to 2P . We find
the values of the first 30 singular values σi of the matrix Ỹrelax for systems
with different P .
−−→
Figure 8.10 shows σi , i = 1, 30 in the deterministic case εY = 0. A
sharp jump in values clearly indicates the true dimension of the system.
Now we consider the case with the presence of uncertainty. Figure 8.11
shows the same entities when εY = 10−3 .
The gap in values has become much smaller, but can also be recognized.
Let us now consider how the magnitude of the first “noise” singular num-
ber σ2P +1 relates to the threshold value defined in (8.16). Since we set the
measurement noise level as a relative value εY , we must use normalization.
We consider the ratio of singular values
 N  √ √
σ2P +1 Y  2 Rl · εY · Y max 2 Rl · εY · Y max
= 
 ≤
relax F




 ≤ 


 .
σ1 Ỹrelax  Ỹrelax  Ỹrelax 
F F max
190 Identification of Complex Systems in the Class

Figure 8.10 First singular values of the matrix Yrelax for systems of different order

Figure 8.11 First singular values of the matrix Ỹrelax for systems of different order. Case of
relative error bound εY = 10−3

  √
  σ +1
Let us put that Y max ≈ Ỹrelax  , then 2Pσ 1
≤ 2 Rl · εY . Therefore,
max
let’s build a graph of the singular values normalized to the first of them.
Figure 8.12 shows normalized singular values for εY = 10−3 and a threshold.
 8.4 Simulation Results 191

Figure 8.12 First normalized singular values of the matrix Ỹrelax for systems of different
order. Case of relative error bound εY = 10−3

√
As we can see, the guaranteed estimate 2 Rl · εY is overestimated due
to the ratio of matrix norms  · F and  · max . To get rid of this we
check the possibility of determining the dimension directly in max norm.
Yrelax
N (n̂)max
Figure 8.13 shows values σ̃n̂+1 = for εY = 10−3 and a
Ỹrelax max
threshold 2εY .

N
Figure 8.13 Relative max norm of the matrix Yrelax
192 Identification of Complex Systems in the Class

The threshold 2εY is much more accurate, but values σ̃n̂+1 are not
monotonically decreasing. As we can see, this method can also be used to
determine the dimension.
Thanks to a good selection of input signals, it is possible to accurately
determine the order of the system up to high dimensions.

8.4.7 Maximum Dimension of Identifiable Model

Despite the fairly good possibilities of determining the exact dimension of the
system, this is not a guarantee that the problem of identifying a model of this
dimension will be correctly posed. Let us check the output error of full-size
models identified under conditions of uncertainty in the data.
We use a set of favorable systems, similar to the one used in the previous
paragraph. Identification is carried with next settings: use the signal matrix
S
Yrelax (n) is on, KN S = 100, 50 realizations of noise with a relative bound
εY = 10−10 . Results are collected in a Table 8.4.
As we can see, the model of dimension 22 cannot be considered suitable in
any way. This means that even in the case of determining the exact, but high
dimension of the system, it is necessary to restrict ourselves only to models of
reduced dimension. The limiting dimension, in addition to the characteristics
of the system itself, depends on the magnitude of the noise.
Figure 8.14 shows values dependence of the accuracy of the model on its
dimension and the magnitude of the noise.

Table 8.4 Quality of models found

Relative error by output
P
Maximum Average
2 4.6773e-11 2.1630e-11
3 1.6977e-10 8.7567e-11
4 1.1200e-09 6.0051e-10
5 1.1383e-08 5.3145e-09
6 1.1548e-07 5.2235e-08
7 9.8088e-07 4.7196e-07
8 9.0444e-06 4.0171e-06
9 9.0678e-05 4.4297e-05
10 0.003637 0.002871
11 0.243185 0.228510
 8.5 Conclusion 193

Figure 8.14 Accuracy of the model, depending on the dimension and the noise

8.5 Conclusion
In the classical statistical method, it is assumed that the implemented errors
have the specified statistical properties, which ensure the consistency of the
estimation, i.e. convergence to unbiased estimates of system parameters. In
case of non-stochastic identification, arbitrary random errors in the data,
including unfavorable ones, are allowed. In these cases, in order to ensure
the quality of the models, during active experiments, it is necessary to pay
serious attention to the information content of the system input. In this
chapter, this issue is focused on. It is substantiated how the experiment should
be planned so that the SNR for each of the modes of the system was the
greatest. Numerical modeling has been carried out, which makes it possi-
ble to evaluate the effectiveness of the proposed approach to identification
and interactively construct a mathematical model of the system, taking into
account its structural features and experimental capabilities, on which the
quality of the resulting models significantly depends.
194 Identification of Complex Systems in the Class

References
[1] Yuriy P. Kondratenko, Arkadii A. Chikri, Vyacheslav F. Gubarev, Janusz
Kacprzyk (Eds.), ‘Advanced Control Techniques in Complex Engi-
neering Systems: Theory and Applications’, Dedicated to Professor
Vsevolod M. Kuntsevich, Studies in Systems, Decision and Control, vol.
203, Springer Nature Switzerland AG, 337 pp., 2019.
[2] T.Söderström, & P.Stocia, ‘System identification’ U.K.: Prentice-Hall
International, Hemel Hempstead, 612 pp., 1989.
[3] L.Ljung, ‘System Identification: Theory for the user’, Prentice-Hall,
1999.
[4] T. Söderström, ‘Errors-in-variables methods in system identification’,
Springer, 2018.
[5] Van Huffel, S., & Lemmerling, P. (Eds.), ‘Total least square and errors-
in-variables modeling: Analysis, algorithms and applications’, Springer,
2002.
[6] S. Beghelli, R.P. Guidorzi, U. Soverini, ‘The frisch scheme in dynamic
system identification’, Automatica, vol. 26, pp. 171–176, 1990.
[7] M. Hong, T. Söderström, W.X. Zheng, ‘Accuracy analysis of bias-
eliminating least squares estimates for errors-in-variables systems’,
Automatica, vol. 43, pp. 1590–1596, 2007.
[8] T. Söderström, ‘Accuracy analysis of the Frisch scheme for identifying
errors-in-variables systems’, IEEE Transactions on Automatic Control,
vol. 52, no 6, pp. 985–997, 2007.
[9] T. Söderström, ‘Errors-in-variables methods in system identification’,
Automatica, vol. 43, pp. 939–958, 2007.
[10] T. Söderström, U. Soverini, ‘Errors-in-variables identification using
maximum likelihood estimation in the frequency domain’, Automatica,
vol. 79, pp. 131–143, 2017.
[11] E. Zhang, R. Pintelon, ‘Identification of multivariable dynamic
errors-in-variables system with arbitrary inputs’, Automatica, vol. 82,
pp. 69–78, 2017.
[12] E. Zhang, R. Pintelon, ‘Nonparametric identification of linear dynamic
errors-in-variables systems’, Automatica, vol. 94, pp. 416–425, 2018.
[13] Zheng, W.X. & Feng, C.B., ‘Unbiased parameter estimation of linear
systems in presence of input and output noise’, International Journal of
Adaptive Control and Signal Processing, vol. 3, pp. 231–251, 1989.
[14] Georgiou, T.T., & Smith, M.C., ‘Optimal robustness in the gap metric’,
IEEE Transactions on Automatic Control, 35, pp. 673–686, 1990.
 References 195

[15] Hyundeok Kang, Guoxiang Gu, Wei Xing Zheng, ‘A graph subspace
approach to system identification based on errors-in-variables system
models’, Automatica, Volume 109, 2019, 108535.
[16] Kuntsevich, V.M., Gubarev, V.F., Kondratenko, Y.P., Lebedev, D.V.,
Lysenko, V.P. (Eds), ‘Control Systems: Theory and Applications’. River
Publishers Series in Automation, Control and Robotics, 329 pp., 2018.
[17] Kondratenko, Y.P., Kuntsevich, V.M., Chikrii, A.A., Gubarev, V.F (Eds),
‘Advanced Control Systems: Theory and Applications’, River Publish-
ers Series in Automation, Control and Robotics, 441 p., 2021.
[18] Verhaegen M. and Dewilde P, ‘Subspace model identification. Part 1:
The output-error state space model identification class of algorithms’,
International Journal of Control, Vol. 56, No 5, pp. 1187–1210, 1992.
[19] V.F. Gubarev, ‘Modeling and Identification of Complex Systems’,
Naukova Dumka, 246 p. 2019 (in Ukrainian).
[20] G.H. Golub, Ch.F. Loan, ‘Matrix Computations’, The John Hopkins
University Press, Baltimore, 780 p., 2013.
[21] A.N. Tikhonov and V.Y Arsenin, ‘Solution of ill-posed problems’, V.H.
Winston & Sons, Washington, D.C.: John Wiley & Sons, New York, 258
p., 1977.
 9
Fuzzy Systems Design: Optimal Selection
of Linguistic Terms Number

Oleksiy Kozlov1 , Yuriy Kondratenko1 , Oleksandr Skakodub1 ,

and Zbigniew Gomolka2
1 PetroMohyla Black Sea National University, 10 68th Desantnykiv st.,
Mykolayiv, 54003, Ukraine
2 University of Rzeszow, Department of Computer Engineering,

Rzeszow, Poland
E-mail: [email protected]; [email protected];
[email protected]; [email protected]

Abstract
This chapter considers developing and researching advanced information
technology for fuzzy systems (FSs) design and structural optimization with
optimal linguistic terms (LTs) numbers. The application of the proposed
information technology allows increasing the FS accuracy and performance,
reducing the computational costs spent on the generation of rule base (RB)
and optimization of parameters, and simplifying its software and hardware
implementation. The effectiveness study of the developed information tech-
nology is carried out on a specific example, particularly when designing and
optimizing the fuzzy control system of a quadrotor unmanned aerial vehicle
(QUAV). The research results obtained confirm the high effectiveness of the
proposed information technology and the feasibility of its use in creating
various types of fuzzy decision-making and control systems.

Keywords: A fuzzy system, design, and structural optimization, information

technology, linguistic terms, fuzzy controller, quadrotor unmanned aerial
vehicle.

197
198 Fuzzy Systems Design: Optimal Selection

9.1 Introduction
In today’s world, artificial intelligence (AI) is the engine of overall progress.
AI technologies can help address global challenges such as combating cli-
mate change and optimizing results in transport, medicine, agriculture, etc.
[1, 2]. In particular, modern AI technologies are used for trading financial
instruments on stock exchanges, carrying out crewless vehicles, establish-
ing medical diagnoses, analyzing large amounts of data, recognizing and
generating images, creating industrial and household robots, as well as high-
precision autonomous weapons [3, 4]. AI processes vast amounts of data,
making conclusions with the help of specific algorithms, which become the
basis for independent decision-making.
Fuzzy logic is a relatively widespread and powerful AI tool with great
potential [5–7]. It allows to operate effectively with expert knowledge,
simulate human decision-making processes, and form linguistic models of
complicated dynamic plants [8–10]. The most expedient is the application of
methods and means of fuzzy logic to construct intelligent decision-making
and control systems of different types. In particular, expert systems for
decision-making based on fuzzy logic techniques are successfully used in
conditions of uncertainty in many areas: financial management, technical
and medical diagnostics, stock market forecasting, transport logistics, and
others [11–13]. Also, fuzzy devices and inference systems are developed and
frequently implemented as observers and adaptive devices, controllers and
identifiers, units of tactical and strategic control for automation of complex
nonstationary and nonlinear objects, such as power plants, robotic production
lines, chemical reactors, spacecraft, mobile robots, satellites, marine ships,
and floating structures, crewless underwater vehicles, drones and others
[14–16].
A specific feature of units and systems based on fuzzy logic is a consider-
able number of structural elements and parameters that significantly influence
their efficiency and should be tuned by the experts or automatically using spe-
cific progressive techniques during their design process [17–19]. In particular,
the main adjustable parameters and elements are: (a) the shapes and parame-
ters of the linguistic terms membership functions, (b) the number of LTs for
input and output signals, (c) the types of operations of aggregation, activation,
accumulation, and the defuzzification procedures, (d) the number of rules of
the fuzzy rule base and its antecedents and consequents, (e) the adjustable
input gains, etc. [20–22]. It gives the ability to create effective, flexible,
and complicated strategies of decision-making and control but, at the same
 9.2 Related Works and Problem Statement 199

time, significantly complicates the development process of the given units and
systems. Thus, in many cases, using expert assessments, fuzzy logic systems
do not enormously increase the main quality indicators compared to similar
systems based on conventional principles. Moreover, they may require more
computational costs and more powerful hardware [23–25]. However, recently,
the development and implementation of perspective progressive algorithms,
approaches, and information technologies for fuzzy systems design based on
specific optimization procedures are one of the leading directions of research
in the FSs modern theory [26–28]. Modern fuzzy systems developed using
these advanced optimization-based techniques can be successfully trained
like neural networks, using training samples or objective functions for high
accuracy attainment, and still maintain good interpretability and transparency
[29–31].
This chapter focuses on developing and researching an advanced infor-
mation technology for designing and optimizing fuzzy systems based on the
optimal selection of linguistic terms number. The structure of the paper is
organized as follows. Section 9.2 presents the brief literature survey, the
statement of the research problem, and the primary purpose of this study.
Section 9.3 sets out in detail the proposed information technology of the
FS structural optimization based on the optimal selection of linguistic terms
number. Section 9.4 presents the results of studying the effectiveness of the
developed information technology on a specific example of a fuzzy control
system for the quadrotor unmanned aerial vehicle with a detailed discussion
of the results of computer simulation. Finally, Section 9.5 concludes the work
and suggests the directions for future studies.

9.2 Related Works and Problem Statement

Many papers are currently published devoted to developing and appro-
bating fuzzy control and decision-making systems of different types and
fields [32–34]. Also, many studies focus on the challenges of synthesis and
structural-parametric optimization of fuzzy systems that are solved through
specifically developed methods, algorithms, and information technologies
[35–37]. In particular, approaches and information technologies of design
with optimization of the weight gains of rule bases and linguistic terms
membership functions are discussed in papers [38–40]. In turn, methods
of optimization of fuzzy systems’ structures based on optimal selection of
defuzzification procedures and RB interpolation and reduction are considered
in [41–43]. The cutting-edge studies show that nature-inspired intelligent
200 Fuzzy Systems Design: Optimal Selection

techniques are promising for synthesizing fuzzy systems and structural-

parametric optimization. They have substantial advantages over classical
optimization methods [44–46]. The main of them are: algorithms of ant
colony optimization [47–49], particle swarm optimization [50–52], biogeog-
raphy based optimization [53–55], grey wolf optimization [56–58], genetic
methods [59–61] and evolutionary strategies [62], immune methods [63–65]
and others [66–69]. The main benefits of these intelligent techniques are as
follows: 1) the opportunity of effective optimization of fuzzy systems of large
dimension; 2) the possibility of detailed studying of large, multimodal, and
nonsmooth search space, excluding looping at local minima; 3) absence of
any restrictions on the FS objective functions [70–72].
When analyzing the task of developing and optimizing a fuzzy system
using the given above nature-inspired intelligent methods and information
technologies, it is necessary to pay special attention to the structural opti-
mization issue [39, 73, 74]. The given challenge is the most important since
the optimal variant of the system’s structure found in the search process
allows not only to improve the accuracy, interpretability and transparency.
It also reduces the computational costs spent on the RB creation and para-
metric optimization and the complexity of the FS software and hardware
implementation. The best structure of the fuzzy inference system is such a
variant of structure that provides: (a) attainment of the high accuracy, inter-
pretability and transparency; (b) reasonable computational costs spent on the
RB creation and optimization of parameters; (c) the admissible complexity of
the FS software and hardware implementation [39, 73, 75]. Therefore, stud-
ies directed towards developing and enhancing approaches and information
technologies for obtaining the fuzzy systems’ optimal structures are relevant
and urgent for the advanced theory of fuzzy control and decision-making
systems.
The scope of main tasks of structural optimization includes the following
tasks: determination of the LTs optimal number and membership functions
types; determination of the optimal procedures of fuzzy inference engine
(FIE), which include aggregation, activation and accumulation; determination
of the optimal defuzzification procedure; reduction of the RB antecedents and
number of rules [73, 76, 77]. Herewith, reduction of the rule base antecedents
and number of rules can be performed separately after the generation of
the structure and all the FS parameters to increase its interpretability and
transparency and simplify further software and hardware implementation
[41, 42, 46]. Also, these optimization procedures do not affect FS parametric
optimization and optimization of other structural elements. Moreover, the
 9.2 Related Works and Problem Statement 201

determination of the optimal procedures of a fuzzy inference engine and

defuzzification procedure can also be performed separately after synthesizing
the FS structure and all the parameters. It increases its performance and
accuracy and does not affect FS parametric optimization and optimization
of other structure elements [28, 39]. In turn, the optimization task of the
LT membership functions types can be solved at the determined number
of terms and at a composed rule base for improving FS performance and
accuracy and simplifying its subsequent parametric optimization procedures
[73]. Thus, conducting this optimization procedure significantly affects the
further parametric optimization of the FS, namely, the number of optimized
parameters of LTs, which depends on the shapes of their MF. As for the
problem of determining the optimal number of linguistic terms, the efficiency
of its solution significantly affects the performance and accuracy and the
complexity of the subsequent procedures of parametric optimization and soft-
ware and hardware implementation of the FS. The setting of various numbers
of linguistic terms allows implementing different strategies of control and
decision-making at forming the rule bases of FSs. So, the number of terms
for input and output signals affects the initial antecedents sets, the number of
fuzzy rules and RB possible consequents, and the number of terms parameters
that should be further optimized during the FS development process. In
addition, it is necessary to compose a new rule base to calculate the FS
efficiency for each new variant obtained in the process of the LTs number
optimization, which requires high computational costs. Therefore, selecting
the optimal number of linguistic terms is the most complicated and vital task
of structural optimization of fuzzy systems, the solution of which is the focus
of the present work. For detailed studying of the problem of selecting the LTs
optimal number, next, consider the generalized MIMO fuzzy system with the
specialized structural optimization unit.
The functional structure of the generalized MIMO fuzzy control/decision-
making system with structural optimization unit that implements the optimal
selection of LTs number is presented in Figure 9.1, where the following
notations are accepted: X is the vector of FS inputs; Y is the vector of FS
outputs; S is the vector that determines numbers of LTs for fuzzy system
inputs and outputs; R is the vector that determines RB rules that consists of
the corresponding antecedents and consequents; LT is the vector that deter-
mines certain LTs for FS inputs and outputs, based on which the antecedents
and consequents of the rules are formed; J is the objective function that
evaluates the FS performance and accuracy; Z is the vector of the outputs
202 Fuzzy Systems Design: Optimal Selection

Figure 9.1 Structure of the generalized MIMO fuzzy control/decision-making system with
structural optimization unit

of the control plant/operating environment that characterize its performance

and quality indicators.
The presented generalized fuzzy system (Figure 9.1) implements the
following nonlinear dependence f F S [29, 45]

Y = fFS (X),
(9.1)
Y = (y1 , y2 , ..., yj , ..., ym ), X = (x1 , x2 , ..., xi , ..., xn ),

where X is the vector of FS n input variables x1 , x2 , . . . , xi , . . . , xn ; Y is the

vector of FS m output variables y1 , y2 , . . . , yj , . . . , ym .
The fuzzy inference engine transforms input signals X, performing
sequential operations (fuzzification, aggregation, activation, accumulation,
and defuzzification) to obtain the output variables Y [29, 78, 79]. Spe-
cific sets of linguistic terms (vector LT) are used for fuzzification and
defuzzification of FS inputs and outputs. They depend on the selected LTs
numbers S.
 9.2 Related Works and Problem Statement 203

The calculation of the fuzzy values of the system outputs is performed

based on the set of rules stored in the rule base and correspond to the vector
R. Each fuzzy rule consists of the specific antecedent and consequent and, for
example, can be defined by the expression [29]

IF “x1 = A1 AND “x2 = A2 AND . . . AND

“xi = A3 ... AND . . . AND “xn = A4
THEN “y1 = B1 AND “y2 = B2 AND . . . AND
“yj = B3 ... AND . . . AND “ym = B4 , (9.2)

where A1 , A2 , A3 , A4 , B1 , B2 , B3 and B4 are corresponding LTs for the FS

input and output variables.
To solve the task of selecting the optimal number of linguistic terms, the
given fuzzy system (Figure 9.1) uses the specialized structural optimization
unit that consists of the objective function calculation unit, the unit for LTs
optimal number selection, and the unit that implements the automatic RB
synthesis. In turn, the unit for LTs number optimal selection should provide
finding such a vector S that will allow the FS high performance, accuracy
and quality indicators, and admissible complexity of the FS software and
hardware implementation. In this case, the vector S is presented in the
following form

S = {Sxi , Syj }, i = (1, 2, ..., n), j = (1, 2, ..., m), (9.3)

where Sxi and Syj are selected numbers of LTs for ith input and jth output FS
variables.
For example for fuzzy system with 3 input variables (n = 3) and 2 output
variables (m = 2) the following numbers of LTs are selected for the given
variables: Sx1 = 5, Sx2 = 3, Sx3 = 3, Sy1 = 5, Sy2 = 7. Then, the vector S has
the following form
S = {5, 3, 3, 5, 7}. (9.4)
In turn, for each found variant of the vector S by the unit of LTs number
optimal selection, the new variant of the rule base (vector R) should be syn-
thesized. Only then, the objective function J can be calculated for evaluation
FS with obtained configuration. Thus, the task of the RB synthesis is the sub-
task of the LTs number optimal selection, which should be solved each time
at every iteration during the implementation of the LTs number optimization
process. The given subtask should be solved in automatic mode by the RB
synthesis unit. In addition, in the optimization process, both units of the LTs
204 Fuzzy Systems Design: Optimal Selection

number optimal selection and the RB synthesis use objective function J to

evaluate the FS performance, accuracy, quality indicators, and complexity of
software and hardware implementation. The objective function calculation
unit should be applied to calculate the current value of the objective function
J based on the vector Z of the control plant’s outputs.
Thus for effective solving the presented complicated task of the optimal
selection of the number of the linguistic terms for fuzzy control and decision-
making systems, the given structural optimization unit should implement
the specifically developed advanced information technology based on the
bioinspired intelligent optimization methods.
The primary purpose of this work is the development and research of
advanced information technology for fuzzy systems design and structural
optimization with optimal selection of the number of linguistic terms based
on the bioinspired intelligent methods.

9.3 Information Technology for Fuzzy Systems Design and

Structural Optimization with Optimal Selection of
Linguistic Terms Number
In this work, we developed two modifications of the information technology
for fuzzy systems design and structural optimization with optimal selection of
linguistic terms number: 1) based on the sequential search approach; 2) based
on the random search approach. Moreover, each modification implements
a nature-inspired technique of ant colony optimization (ACO) [29] and a
method of sequential search [21] for automatic development of the rule
base that corresponds to each new generated vector S that defines linguistic
terms number. The first modification based on the sequential search uses the
approach of sequential selection of the best values of the number of the terms
for the FS, starting from the first FS input x1 and ending with its last output
ym . This modification consists of the following successive stages.
Stage 1. Choosing of the FS inputs and outputs. At the given stage, for the
developed FS n inputs and m outputs are chosen. In turn, the total number of
the system’s variables n + m is selected, taking into account the peculiarities
of the specific problem that FS should solve.
Stage 2. The setting of the minimum and maximum values of FS inputs
and outputs. At this stage, the operating ranges between the minimum and
maximum values are set for all n input and m output variables. For instance,
if all signals come to the system’s input in relative units to their maximum
 9.3 Information Technology for Fuzzy Systems Design 205

values, the operating ranges should be set from –1 to 1 [80, 81]. In addition, if
some inputs or outputs can be only positive by nature (electric heating power,
water or gas consumption, and others), then their changing ranges should be
from 0 to 1.
Stage 3. Choosing the shapes of LT membership functions for FS inputs and
outputs. The shapes of linguistic terms MFs for all n input and m output
variables are chosen. Often, the same shapes are selected for all LTMFs of the
system’s variables at the initial stage of the designing process. For instance,
Gaussian, trapezoidal or triangular shapes may be chosen [82, 83]. Also, the
initial values of their parameters for each FS input and output should be
set in an automatic mode so that the LTs would be evenly distributed over
the operating ranges (between the minimum and maximum values) for any
specified LTs number.
Stage 4. Determination of the vector S initial structure. The vector S that
defines LTs numbers for FS input and output variables is formed at this stage,
taking into account the chosen at the Stage 1 n inputs and m outputs. It has
the following form

S = {Sx1 , Sx2 , ..., Sxi , ..., Sxn , Sy1 , Sy2 , ..., Syj , ..., Sym }. (9.5)

The given vector includes variables Sxi,yj that define the following num-
bers of the LTs for the fuzzy system’s variables. These variables are located
in order, starting from the first input x1 and ending with the final output ym .
Stage 5. The setting of vector S constraints. The constraints Smin and Smax
on the LTs number are selected for each FS input and output. For example,
the minimum value of the number of the terms may be selected equal to 2
(Smin = 2), and the maximum value can be chosen equal to 7 (Smax = 7)
for each FS input [28, 39]. In turn, for the FS outputs, the minimum and
maximum values can be equal to 3 (Smin = 3) and 9 (Smax = 9), respectively
[82]. Moreover, at this stage, some additional constraints may be selected. For
instance, for fuzzy control systems, if their variables can vary from positive
to negative values within the symmetric operating ranges, the additional
constraints may be set, that the number of the linguistic terms should be only
odd [39]. Therefore, in this case, in addition to basic constraints (Smin and
Smax ), the LTs number maybe only be as follows: 3, 5, and 7 – for FS inputs
(Sxi ∈ {3, 5, 7}); 3, 5, 7, and 9 – for output variables (Syj ∈ {3, 5, 7, 9}).
Stage 6. Selection of the FS complex objective function J C . The main
parameters and boundary (optimal) value of the complex objective function
206 Fuzzy Systems Design: Optimal Selection

J C for evaluating the effectiveness of the developed FS with a certain number

of LTs are selected at this stage. As the number of rules, the antecedents
sets, and possible consequents of the system’s RB directly depend on the
determined number of linguistic terms, then two criteria should be used in
the vector S structural optimization process. The criterion J 1 , which evaluates
the developed control or decision-making FS effectiveness, and criterion J 2 ,
which considers the complexity of the designed rule base and its software
and hardware realization [73]. Therefore, the problem of the optimal LTs
number determining is the multi-criteria optimization problem [73, 84, 85],
solve which it is necessary to find a compromise minimum of two objective
functions J 1 and J 2 . At solving this problem, the prior approach for multi-
criteria optimization with aggregation of several objective functions is most
advisable [73, 85]. According to this approach, two criteria J 1 and J 2 , with the
initial evaluation of their importance, are combined into one global criterion
the complex objective function J C . Thus, the current value of the complex
objective function J C should be calculated during the search process based
on the following equation

JC = J1 + kJ2 J2 , (9.6)

where kJ 2 is the scale coefficient at J 2 that defines the significance of

considering this objective function during the optimization process of the
number of linguistic terms of the fuzzy system.
For example, suppose the fuzzy system is designed based on a training
sample. In that case, the mean squared error (MSE) may be used as the
criterion J 1 , which measures the average of the squares of the errors between
the estimated and actual FS outputs. In the case of the fuzzy control system,
the objective function J 1 can be selected as an integral quadratic error
of control or the integral quadratic deviation between the outputs of the
reference model and the actual control plant [29, 30, 86]. In turn, for the fuzzy
decision-making system, the criterion J 1 may be calculated as the percentage
of wrong decisions obtained for the test dataset.
As for the values of the objective function J 2 , which evaluates the com-
plexity of the designed rule base as well as software and hardware realization
of the fuzzy system, they should be calculated based on the expression [45]
⎛ ⎞ 1 2⎛ m ⎞
!m =n !
J2 = rmax ⎝ Syj ⎠ = Sxi ⎝ Syj ⎠ , (9.7)
j=1 i=1 j=1
 9.3 Information Technology for Fuzzy Systems Design 207

where rmax is the total number of the RB rules, defined by the number of all
possible combinations of the LTs of the system’s inputs.
Moreover, the boundary (optimal) values of the functions J 1opt and J Copt
and the corresponding coefficient kJ 2 are preliminarily selected at this stage.
It is necessary to consider the main requirements and peculiarities of the
problem of fuzzy system design. For example, the optimal value of the
objective function J Copt and the scaling factor kJ 2 can be selected based on
the following conditions
JCopt JCopt JCopt
< 2, > 2, = 2. (9.8)
J1opt J1opt J1opt

Stage 7. Determination of the vector S initial value. The initial values of

LTs numbers (vector S0 ) of the FS inputs and outputs are defined at this
stage, which are the starting points for the optimization process. In turn, the
vector S0 should be set as its minimum value Smin that can be determined as
constraints at Stage 5

S0 = Smin = {Sx1 min , ..., Sxi min , ..., Sxn min ,

Sy1 min , ..., Syj min , ..., Sym min }. (9.9)

At the following stages of the proposed information technology, the

sequential search of the numbers of linguistic terms of the FS is performed,
starting from the first FS input x1 and ending with its last output ym .
Stage 8. Selection of the first FS variable. At this stage, the transition to the
first input variable x1 of the fuzzy system is conducted to start the iterative
sequential search for obtaining the optimal vector Sopt of the number of
linguistic terms. Furthermore, the iterative search procedure starts from the
initial value S0 , previously determined at Stage 7.
Stage 9. Checklist check. Each variant of vector S, for which the fuzzy system
RB is already developed with calculating the complex objective function
J C current value in the given information technology implementation, is
recorded to the previously created Checklist. Also, the corresponding rule
base and value of the system’s objective function J C enter into the Checklist.
At this stage, the Checklist check is implemented to avoid the repeated rule
base generation and additional calculations of its objective function J C for
the system with the same vector S of linguistic terms number. It gives the
opportunity not to perform unnecessary iterations with rule base development
and calculation of the complex objective function J C , which number is n +
208 Fuzzy Systems Design: Optimal Selection

m –1. If the current vector of the LTs number S is already recorded to the
Checklist, then transition to Stage 13 is performed. In the opposite case, go
to Stage 10.
Stage 10. Development of the rule base for the system with the current vector
S and calculation of its complex objective function. The rule base generation
is conducted at this stage through the ACO-based method or technique of
the sequential search, designed and approbated in papers [29] and [21],
respectively. Also, the complex objective function J C value is calculated
for this designed rule base. As the previous comprehensive research [45]
shows, it is expedient to choose the sequential search technique or the
nature-inspired ACO-based method for performing an effective automatic
RB design, depending on its size and complexity, defined with the help of
criterion J 2 (9.7). Herewith, if the criterion J 2 ≤ 600, the sequential search
approach should be used. If J 2 > 600, then the nature-inspired ACO-based
technique will be more efficacious [45]. Consequently, at this stage, the
criterion J 2 (9.7) for the current vector S is first calculated, and only then,
based on its defined value, a suitable method for RB design is chosen (ACO-
based method or sequential search). Further, the rule base development is
performed with the calculation of the criterion J 1 using the chosen technique.
After that, the value of the complex objective function J C is finally calculated
by the equation (9.6) using values of the criteria J 1 and J 2 . In turn, the
adjustable parameters of the ACO-based and sequential search algorithms
(number of rounds of sequential search l, number of agents in the population
Z max , number of elite agents e, maximum number of iterations N*max , etc.)
are previously set based on conducted experimental studies and obtained
recommendations in the previous research [29, 45]. The antecedents of the
rule base for the current vector S are synthesized in an automatic mode as all
possible combinations of LTs of the system’s inputs. As for their consequents,
they are determined during the design process through the above sequential
search method or ACO-based approach.
Stage 11. Check for the attainment of the complex objective function optimal
value. At this stage, the achievement of the optimal value of the complex
objective function J Copt is checked for the obtained vector S of the LTs
number. If the given check is positive, then transition to Stage 16 is carried
out. Otherwise, go to Stage 12.
Stage 12. Record the current vector S, the corresponding rule base, and the
system’s value of the complex objective function to the Checklist. The current
vector S, the corresponding developed rule base, and the system’s current
 9.3 Information Technology for Fuzzy Systems Design 209

value of the complex objective function J C are recorded to the Checklist at

this stage.
Stage 13. Check the optimization process completion for the FS current
variable. Optimization calculations are considered complete for each FS
variable if the RBs were developed for every existing variant of the number
of the linguistic terms Sxi,yj within the set constraints [Sxi,yj min , Sxi,yj max ]
for the given variable. If this check is positive, then transition to Stage 14 is
carried out. In the opposite case, one is added to the current LTs number Sxi,yj
for the given variable (Sxi,yj + 1), and then return to Stage 9 is performed.
Stage 14. Choosing the best value of the number of linguistic terms for the
current system’s variable. Determination of the number of terms Sxi,yj best
is performed at this stage, for which the complex objective function J C is
the smallest (J Cbest ) among all generated during the search process for this
specific variable. Then, this variant of the LTs number Sxi,yj best is set for the
given variable.
Stage 15. Check the optimization process completion for all the system’s
variables. The calculations of structural optimization are considered complete
at this stage if the linguistic terms numbers were optimized and the best
variants Sbest were selected and set for all n + m variables (starting with
the first input x1 and ending with the last output ym ). If the given check is
positive, then the transition to Stage 16 is performed. Otherwise, go to the
next FS variable (i, j + 1) and return to Stage 9.
Stage 16. Completion of the optimization process of the LTs number. When
the LTs number optimization process is complete, the additional parametric
or structural optimization of the FS membership functions can be carried out
and the system’s implementation using proper hardware and software [87].
In turn, the software and hardware realization will be significantly simplified
due to the obtained optimal structure of the fuzzy system and the optimal
number of rules of its rule base.
Figure 9.2 shows a flowchart of the first modification of the pro-
posed information technology for fuzzy systems development and structural
optimization with optimal selection of LTs number.
At the implementation of the given modification of the proposed infor-
mation technology, the maximum number of iterations N max is determined
based on the expression
!
n !
m
Nmax = (ΔSxi + 1) + (ΔSyj + 1) − (n + m − 1) , (9.10)
i=1 j=1
210 Fuzzy Systems Design: Optimal Selection

Figure 9.2 Flowchart of the first modification of the information technology for FS structural
optimization based on optimal selection of linguistic terms number

where
ΔSxi = Sxi max − Sxi min , i = 1, 2, ..., n; (9.11)
ΔSyj = Syj max − Syj min , j = 1, 2, ..., m. (9.12)
Moreover, each N-th iteration of the presented information technology
includes conducting N*max iterations of the ACO-based method or sequential
search technique at Stage 10.
The second modification based on the random search approach uses the
concept of creating H random variants of vectors of the LTs number and
 9.3 Information Technology for Fuzzy Systems Design 211

choosing the best from them. The first six stages of this modification are
the same as for the first modification of the information technology. Next,
consider the rest of the stages of this modification, starting from Stage 7.
Stage 7. Generation of H different variants of vector S in a random way.
At this stage, H different variants of the vector of LTs number are randomly
generated. In turn, each variable Sxi,yj of each h-th vector Sh is generated ran-
domly within the constraints Sxi,yj min and Sxi,yj max , that are set at Sage 5.
Moreover, the number of random variants H is previously selected before
implementing this modification of the information technology. Herewith, if
there are some identical vectors among the generated variants, then they are
deleted, and new ones are generated in their place until all H vectors S are
different. The rule bases will be synthesized, and the complex objective func-
tion J C values will be calculated for each h-th vector of the given generated
vectors at the further stages of the given modification of the information
technology.
Stage 8. Selection of the first variant of the generated vectors S. The first
variant (h = 1) of the generated vectors of the LTs number S1 is selected at
this stage to start the procedures of the optimization process.
Stage 9. Development of the rule base for the system with the current vector
S and calculation of its complex objective function. The given stage of
this modification is the same as Stage 10 for the first modification of the
information technology.
Stage 10. Check for the attainment of the complex objective function optimal
value. The given stage of this modification is the same as Stage 11 for the first
modification of the information technology. If this check is positive, then the
transition to Stage 13 is carried out. In the opposite case, go to Stage 11.
Stage 11. Check for the optimization process completion of the LTs number
for the FS. The calculations of structural optimization are considered to be
complete at this stage if the design of the rule base and determination of
the complex objective function J C is conducted for all H vectors of the LTs
numbers generated at Stage 7 (h = H). If the given check is positive, then
the transition to Stage 12 is performed. Otherwise, go to the next (h + 1)
generated vector Sh+1 and return to Stage 9.
Stage 12. Selection of the best variant of the generated vectors S. At this
stage, the best variant of the H generated vectors S is selected based on their
values of the complex objective function J C calculated at Stage 9.
212 Fuzzy Systems Design: Optimal Selection

Stage 13. Completion of the LTs number optimization process. The given
stage of this modification is the same as Stage 16 for the first modification of
the information technology.
Figure 9.3 shows a flowchart of the second modification of the pro-
posed information technology for fuzzy systems development and structural
optimization with optimal selection of LTs number.
At the implementation of the given modification of the proposed infor-
mation technology, it is advisable to set the number of generated vectors H
the same as the maximum number of iterations N max for the first modifica-
tion for the objectivity of comparing the computational costs of these two
modifications.

Figure 9.3 Flowchart of the second modification of the information technology for FS
structural optimization based on optimal selection of linguistic terms number
 9.3 Information Technology for Fuzzy Systems Design 213

Figure 9.4 Flowchart of the stage of design of the RB and calculation of the complex
objective function for the FS with the current vector S

Moreover, the flowchart of the stage of design of the RB and deter-

mination of the complex objective function for the fuzzy system with the
current vector S (Stage 10 of the first modification, Stage 9 of the second
modification) for both considered modifications has the form presented in
Figure 9.4.
The main features of the sequential search and ACO-based methods for
automatic rule base synthesis of different types of fuzzy systems are presented
in works [21, 29, 45]. In turn, the sequential search method allows to carry
out synthesis and optimization of highly efficient rule bases with an optimal
set of consequents Ropt , at which the value of the objective function J 1 of
control or decision-making processes based on the developed FS will be
optimal [21]. Antecedents of rules are formed as all possible combinations of
linguistic terms of input variables of FS. The formation of the optimal vector
of the RB consequents Ropt is carried out using an iterative search based on
sequential enumeration of the consequents of each RB rule and calculation of
the corresponding values of the objective function J 1 .
214 Fuzzy Systems Design: Optimal Selection

As for the ACO-based method for automatic rule base synthesis, it is

based on the principles of ant colony optimization and allows to represent
the structure of the fuzzy system RB as a specific graph, on the edges and
nodes of which the agents-ants will move [29].
The route of every Z-th agent of the colony runs through specific edges
and nodes of the given graph at every separate iteration with number N*
[29, 88, 89]. After each Z-th agent passes through the whole graph at every
N*-th iteration, the corresponding RB of fuzzy system with a certain conse-
quents vector RZ (N*) (Z = 1, ..., Zmax ; N ∗ = 1, ..., Nmax∗ ) is formed. The

route length that a specific ant has traveled in the current iteration depends on
the value of the FS objective function (first criterion) J1Z (N ∗).
During the movement of all ants along the graph, the probabil-
ity PrγjZ (N ∗) of the Zth agent passing along the rγjth edge from

the node with number r(j–1) to the rj-th node (r ∈ {1, ..., s} ; γ ∈
{1, ..., Syj } ; j ∈ {1, ..., m}) in the N*-th iteration (N* = 1, . . . , N max ) is
calculated based on the equation [88]

[τrγj (N ∗ )]α [ηrγj ]β

Z
Prγj (N ∗ ) = rvj , (9.13)
∗ α β
rγj=r1j [τrγj (N )] [ηrγj ]

where r is the current rule number of the RB; γ is the current consequent
number of the jth output variable; τrγj (N*) is the pheromone intensity on the
rγjth edge at the iteration N*; ηrγj is an inverse of the relative length Drγj of
the rγjth edge; α is an adjustable parameter which determines the importance
of the pheromone trace considering on edge; β is the parameter that regulates
importance of the relative length Drγj of the rγjth edge.
Other equations and main steps of the ACO-based method and sequential
search method for automatic rule base design of fuzzy control and decision-
making systems are considered in detail in [29, 89].
The effectiveness study of the developed information technology is per-
formed in this work on a specific example, particularly when designing the
fuzzy control system for the quadrotor unmanned aerial vehicle.

9.4 Design and Structural Optimization of the Fuzzy

Control System for the Quadrotor Unmanned Aerial
Vehicle
Quadcopters belong to the micro-class crewless aerial vehicles, which have
recently become very popular and are widely used to implement monitoring
 9.4 Design and Structural Optimization of the Fuzzy Control System 215

and inspection tasks in various areas of the civil and military sector [90–92].
QUAVs have many advantages over other types of crewless aerial vehicles,
namely: high maneuverability, high efficiency, the ability to take off and land
vertically and hovering in space, ease of maintenance, and others. However,
quadcopters are complicated control plants, for automation of which it is
advisable to use intelligent automatic control systems, particularly fuzzy con-
trol systems [93–95]. The stabilization and automatic control of the QUAV
flight altitude when moving over terrain with a complex relief is one of the
most important tasks in automating its spatial motion. In this work, the fuzzy
control system of the QUAV flight altitude is designed to test the effectiveness
of the proposed information technology for FSs structural optimization with
optimal selection of linguistic terms number.
The dynamics of spatial motion of the quadrotor unmanned aerial vehicle
is described by the following system of basic equations [90]:
⎧ 
⎪
⎪ ẍ = (F1 + F2 + F3 + F4 ) (cos φsinθcosψ + sinφ sin ψ)
⎪
⎪ 
⎪
⎪
⎪
⎪ −kx1 (ẋ)2 − kx2 Fx /m;
⎪
⎪ 
⎪
⎪
⎪
⎪ ÿ = (F1 + F2 + F3 + F4 ) (sin φsinθcosψ + cosφ sin ψ)
⎪
⎪ 
⎪
⎪
⎪
⎪ −k ( ẏ) 2
− k F
y2 y /m;
⎪
⎨ 
y1

z̈ = (F1 + F2 + F3 + F4 ) cos θcosψ − Fg − kz1 (ż)2 − kz2 Fz /m;
⎪
⎪   2 
⎪
⎪
⎪
⎪ ψ̈ = l −F1 + F2 + F3 − F4 − kψ1 ψ̇ − kψ2 Fψ /Iψ ;
⎪
⎪
⎪
⎪   2 
⎪
⎪
⎪
⎪ θ̈ = l −F1 − F2 + F3 + F4 − kθ1 θ̇ − kθ2 Fθ /Iθ ;
⎪
⎪
⎪
⎪   2 
⎪
⎪
⎪
⎩ φ̈ = Cφ F1 − F2 + F3 − F4 − kφ1 φ̇ − kφ2 Fφ /Iφ ,
(9.14)
where x, y and z are QUAV’s longitudinal-horizontal and transverse-
horizontal coordinates, as well as flight altitude; ψ, θ and ϕ are the roll,
pitch and yaw angles; m is the QUAV mass; F 1 , F 2 , F 3 and F 4 are the
corresponding values of the rotors’ lifting force; Fx , Fy , Fz , F ψ , F θ, and
F ϕ are the corresponding values of disturbing effects of wind; Fg is the
gravity force; I ψ , I,θ and I ϕ are the QUAV’s moments of inertia about the
longitudinal, transverse and vertical axes; l is the distance from the rotor to
the center of mass; Cϕ is the yaw moment coefficient; kx1 , kx2 , ky1 , ky2 , kz 1 ,
kz 2 , kψ1 , kψ2 , kθ1 , kθ2 , kϕ1 and kϕ2 are the QUAV’s model coefficients.
216 Fuzzy Systems Design: Optimal Selection

Since the time constants of the quadcopter drives are significantly less
than the time constants of the quadcopter itself, then as an assumption,
the dynamics of the rotor drives can be neglected. Thus, the control of the
quadcopter is carried out using four control forces: u1 , u2 , u3 , and u4 [90]. In
turn, these forces are calculated by the equations

u1 = F1 + F2 + F3 + F4 ;
u2 = −F1 − F2 + F3 + F4 ;
(9.15)
u3 = −F1 + F2 + F3 − F4 ;
u4 = F1 − F2 + F3 − F4 .

The development and structural optimization with optimal selection of LTs

number is conducted for the QUAV with the following main parameters:
QUAV’s total mass is 0.32 kg, distance from the rotor to center of mass is
0.209 m, total maximum lift force of 4 propellers is 6 N.
Figure 9.5 presents the functional structure of the FS for control of the
QUAV flight altitude, where the following abbreviations are accepted: CSTL
is the QUAV control system of tactical-level; CSHC is the control system of
horizontal coordinates; SD is the setting device; AS is the altitude sensor,
which directly determines the distance of the quadcopter from the terrain
surface; R1, R2, R3, and R4 are the QUAV’s 1st, 2nd, 3rd and 4th rotors;
zS and zR are the flight altitude set and actual values relative to the terrain
surface; zT and zRT are the absolute altitude values of the terrain surface and
the QUAV, respectively; uSD and uAS are the SD and AS output signals; εz
is the altitude control error of the QUAV flight.
The given control system (Figure 9.5) uses the Mamdani-type fuzzy
controller (FC) for stabilizing and automatic control of the QUAV flight
altitude, 5which inputs are: altitude control error εz , its derivative dε
dt and
z

integral εz dt. In turn, the FC output is u1 . For a more objective study of

the efficiency of the designed fuzzy control system, in this work, the process
of QUAV’s flight at a fixed altitude over mountainous terrain with a complex
relief is considered. In this case, the quadcopter moves at a set constant speed
ẋs along the longitudinal-horizontal axis (x coordinate) over the mountainous
terrain. Herewith, the value of the y coordinate remains unchanged, and the
task of the fuzzy control system of the altitude is to keep the QUAV at a
fixed distance from the surface zS , bypassing around all the unevenness of
the mountainous terrain. In this work, for simulating, we took a 1-kilometer
long section with a complex relief that is randomly generated [96] and is
presented by the dependence zT (x) and shown in Figure 9.6.
 9.4 Design and Structural Optimization of the Fuzzy Control System 217

Figure 9.5 The fuzzy control system of the QUAV flight altitude

Figure 9.6 Complex relief of the mountainous terrain presented by the dependence zT (x)

Then, we conducted the development and structural optimization with

optimal selection of LTs number of the FC alternately using first and second
modifications of the proposed information technology. First, the first six
stages of information technology are performed, which are the same for both
modifications.
218 Fuzzy Systems Design: Optimal Selection

At the first stage, we chose three input variables

5 and one output variable
for the designed fuzzy controller: εz , dε
dt
z
, ε z dt, u1 . Therefore, the total
number of the controller’s variables is equal to 4. At Stage 2, the FC inputs
and output operating ranges are determined from –1 to 1. Then, at the third
stage, triangular shapes of the MFs are chosen for all LTs of the controller’s
inputs and output. The values of LTMF parameters are automatically set to
be evenly distributed over the operating ranges for any specified number for
each input and controller at the further optimization stages. At Stage 4, the
vector S, which defines the numbers of LTs for the controller input and output
variables, is formed in the following way
S = {Sx1 , Sx2 , Sx3 , Sy1 }. (9.16)
At Stage 5, the vector S constraints on the LTs number are determined in
the following way
Sxi ∈ {3, 5, 7} , i = 1, 2, 3;
(9.17)
Sy1 ∈ {5, 7, 9} .
Furthermore, for all input and output variables of the fuzzy altitude con-
troller using 3, 5, 7, and 9 terms, the following sets of LT with the triangular
shape are implemented:
{N, Z, P} ;
{BN, SN, Z, SP, BP} ;
{BN, N, SN, Z, SP, P, BP} ;
{VBN, BN, N, SN, Z, SP, P, BP, VBP} ,
where VBN is very big negative; BN is big negative; N is negative; SN is
small negative; Z is zero; SP is small positive; P is positive; BP is big positive;
VBP is very big positive.
At the sixth stage, the complex objective function J C for evaluating the
efficiency of the synthesized QUAV’s fuzzy control system is determined,
which is calculated by the expression (9.6). In turn, the criterion J 1 is the
mean integral quadratic deviation between the desired motion trajectory zD (x)
and the actual motion trajectory of the QUAV zR (x, ẋs , S)
 xmax ' (
1
J1 (x, ẋs , S) = (zD (x) − zR (x, ẋs , S))2 dx, (9.18)
xmax 0
where xmax is the maximum flight length along the x axis; the desired motion
trajectory zD (x) is as follows
zD (x) = zS + zT (x). (9.19)
 9.4 Design and Structural Optimization of the Fuzzy Control System 219

Moreover, the criterion J 2 is presented by the equation (9.7). As the

optimal values of the criteria J C and J 1 the given values are chosen: J Copt =
0.3; J 1opt = 0.15. The scaling factor kJ 2 for this function is equal to 0.00033.
Then, the seventh stage of the first modification of the proposed informa-
tion technology is conducted, at which the initial value of the vector S for the
fuzzy controller is selected. In this case, the initial value of the vector S is
equal to its minimum possible value Smin according to constraints set at the
fifth stage
S0 = Smin = {3, 3, 3, 5}. (9.20)
Then, the iterative sequential optimization procedure starts with the first
input variable εz and ends with the output variable u1 , according to all the
remaining stages of the first modification (from 8th to 16th). Herewith, the
adjustable parameters of the ACO-based technique and sequential search
method for the rule base design at Stage 10 are previously selected in the
following way. The number of rounds of sequential search l equals 3 for
each possible vector S of the number of the linguistic terms at performing
the sequential search algorithm. As for the ACO-based the population e =
10, the maximum number of iterations N*max = 100, the values of the other
adjustable parameters are as follows: α = 2; β = 1; Q = 0.1; ρ = 0.5.
When conducting the given modification of the proposed information
technology, the maximum number of iterations N max equals 9 (determined
based on the equation (9.10) and selected constraints).
Table 9.1 presents the obtained results (by the first modification) during
the vector S structural optimization for the QUAV altitude fuzzy controller.
Figure 9.7 shows the changing curves of the complex objective function
J C and separately the criteria J 1 and J 2 at implementing the first modification

Table 9.1 Optimization results using the first modification

Iteration N Vector S Criterion J1 Criterion J2 Complex Objective Function JC
1 {3, 3, 3, 5} 0.671 135 0.715
2 {5, 3, 3, 5} 0.249 225 0.323
3 {7, 3, 3, 5} 0.208 315 0.311
4 {5, 5, 3, 5} 0.217 375 0.340
5 {5, 7, 3, 5} 0.202 525 0.375
6 {5, 3, 5, 5} 0.231 375 0.354
7 {5, 3, 7, 5} 0.214 525 0.387
8 {5, 3, 3, 7} 0.119 315 0.223
9 {5, 3, 3, 9} 0.103 405 0.236
220 Fuzzy Systems Design: Optimal Selection

Figure 9.7 Changing curves of the complex objective function J C and the criteria J 1 ,
J 2 values during the structural optimization process (first modification)

of the information technology during the structural optimization process of

the QUAV’s altitude FC.
As can be seen from Table 9.1 and Figure 9.7, the optimal value of the
complex objective function has been achieved (J C ≤ J Copt ) at the eighth
iteration of the given information technology (first modification). Herewith,
the optimization process could be completed according to Stage 11. However,
to conduct a complete study, calculations were carried out for the ninth
iteration as well. In this case, the obtained at the eighth iteration optimal
vector of the number of the linguistic terms has the following form

Sopt = {5, 3, 3, 7} .

In turn, the vector S, obtained at the ninth iteration, has less value of the
criterion J 1 (J 1 = 0.103) than Sopt . However, its criterion J 2 is significantly
bigger (J 2 = 405), which ultimately gives a bigger value of the complex
objective function J C (J C = 0.236). In addition, for the vector S obtained
at the first iteration, the value of criterion J 2 is the smallest (J 2 = 135), but its
criterion J 1 has the worst value (J 1 = 0.671), which as a result gives the worst
value of the complex objective function J C (J C = 0.715).
 9.4 Design and Structural Optimization of the Fuzzy Control System 221

Table 9.2 RB fragment for the vector Sopt obtained using the first modification
Input and Output Variables
Rule Number 5
εz dεz
dt
εz dt u1
1 BN N N BN
5 BN Z Z BN
12 SN N P N
18 SN P P SN
23 Z Z Z Z
27 Z P P P
30 SP N P SP
36 SP P P BP
41 BP Z Z BP
45 BP P P BP

Also, a particular feature of these calculations (carried out using the first
modification) is that the controller RB was developed using the sequential
search technique without using the ACO-based method during all iterations.
Since, for all obtained values of the vector S, criterion J 2 was less than 600.
Table 9.2 presents the RB fragment developed during the optimization
process for the found optimal vector Sopt . The given RB has 45 rules.
Also, the entire vector R of the RB consequents has the following form:
R = (BN, BN, BN, BN, BN, BN, BN, BN, BN, BN, BN, N, N, N, SN, SN,
SN, SN, N, N, SN, SN, Z, Z, SP, SP, P, SP, SP, SP, SP, P, P, SP, P, BP, BP, BP,
BP, BP, BP, BP, BP, BP, BP).
Further, the remaining stages of the second modification of the proposed
information technology are carried out (from 7th to 13th ).
At Stage 7 of the second modification, H different variants of vector S
of LTs number are randomly generated, considering constraints selected at
Stage 5. In this case, the number of generated random variants H is equal
to 9 (H = 9) to be the same as the maximum number of iterations N max for
the first modification for the objectivity of comparing the computational costs
of these two modifications. The rule bases are synthesized, and the complex
objective function J C values are calculated for each h-th vector Sh (h = 1,
2, . . . , H) of the given generated vectors at the further stages of the given
modification. In turn, the RBs are developed through sequential search and
ACO-based methods, depending on the current value of the criterion J 2 . The
authors set the same adjustable parameters of the sequential search method
and ACO-based technique as for the first modification.
222 Fuzzy Systems Design: Optimal Selection

Table 9.3 Optimization results using the second modification

Iteration N Vector S Criterion J1 Criterion J2 Complex Objective Function JC
1 {7, 5, 7, 7} 0.087 1715 0.653
2 {3, 7, 7, 5} 0.341 735 0.583
3 {7, 7, 3, 9} 0.08 1323 0.516
4 {5, 7, 3, 9} 0.095 945 0.407
5 {7, 5, 3, 9} 0.084 945 0.396
6 {3, 3, 7, 9} 0.237 567 0.424
7 {3, 5, 5, 7} 0.203 525 0.376
8 {5, 3, 7, 9} 0.098 945 0.409
9 {7, 7, 5, 5} 0.196 1225 0.601

Table 9.3 presents obtained results during the optimization of the vector
S for the QUAV altitude fuzzy controller using the second modification.
Figure 9.8 shows the changing curves of the complex objective function
J C and separately the criteria J 1 and J 2 at implementing the second modifica-
tion of the information technology during the structural optimization process
of the QUAV’s altitude FC.
As shown from Table 9.3 and Figure 9.8, the optimal value of the complex
objective function J Copt has not been achieved for any generated variant dur-
ing the implementation of the given (second) modification of the information
technology. The best variant (7th variant) of the nine generated vectors is
selected that has the lowest value of the complex objective function (J C =
0.376). It has the following form
Sbest = {3, 5, 5, 7} .
Also, for most of the generated variants of vector S, the synthesis of the
rule base was carried out using the ACO-based approach since the value of
criterion J 2 was more than 600. In turn, the sequential search method was
used only for the sixth and seventh variants.
In general, with the help of this modification, more complex configura-
tions of fuzzy systems with a larger number of LTs for inputs and output were
found in comparison with the variants obtained using the first modification of
the information technology.
The found more complex variants made it possible to achieve lower values
of criterion J 1 (especially 1, 3, 4, 5, 8). However, they have sufficiently
large values of criterion J 2 , which gave unsatisfactory values of the complex
objective function J C for all obtained vectors S. We can assume that with
an increase in the number H of randomly generated variants of vectors of
the number of the linguistic terms, it would be possible to find the optimal
 9.4 Design and Structural Optimization of the Fuzzy Control System 223

Figure 9.8 Changing curves of the complex objective function J C and the criteria J 1 ,
J 2 values during the structural optimization process (second modification)

vector Sopt (with the optimal value of the complex objective function J Copt )
using this second modification. Nevertheless, this would require more com-
putational costs in comparison with the first modification of the information
technology.
Thus, the studies carried out show that the first modification, in general,
has a higher efficiency than the second since it allows to find the optimal
configurations of fuzzy systems (by the LTs number of input and output
variables) using lower computational costs. Also, we can conclude that when
creating fuzzy systems, for which simplicity of configuration, the small size
of the rule base, and simplicity of software and hardware implementation are
essential, it is more expedient to use the first modification of this information
technology. Since it allows synthesizing the simplest and enough effective
fuzzy systems, and, at the same time, uses low computational costs. In turn,
the second modification can be quite effectively used to design fuzzy systems,
for which high accuracy and efficiency of solving the assigned problem are
paramount, and the configuration complexity, the RB size, and the complexity
of the software and hardware implementation are of less importance.
224 Fuzzy Systems Design: Optimal Selection

Table 9.4 RB fragment for the vector Sbest obtained by the second modification
Input and Output Variables
Rule Number 5
εz dεz
dt
εz dt u1
1 N BN BN BN
12 N Z SN BN
20 N SP BP BN
28 Z BN Z SN
38 Z Z Z Z
46 Z BP BN SP
54 P BN SP BP
61 P Z BN BP
68 P SP Z BP
75 P BP BP BP

The RB fragment developed in the optimization process for the obtained

best variant of the vector of LT number Sbest is presented in Table 9.4. The
given RB has 75 rules.
Furthermore, the entire vector R of the RB consequents has the following
form:
R = (BN, BN, BN, BN, BN, BN, BN, BN, BN, BN, BN, BN, BN, BN, BN,
BN, BN, BN, BN, BN, BN, BN, BN, BN, BN, N, N, SN, SN, Z, SN, SN, SN,
SN, Z, SN, SN, Z, SP, SP, Z, SP, SP, SP, P, SP, P, P, BP, BP, BP, BP, BP, BP,
BP, BP, BP, BP, BP, BP, BP, BP, BP, BP, BP, BP, BP, BP, BP, BP, BP, BP, BP,
BP, BP).
Comparing two rule bases developed for vectors Sopt and Sbest obtained
using the first and second modifications, it is evident that for the vector
Sopt the rule base has fewer rules and better value of the complex objective
function J C and better value of criterion J 1 . It suggests that the correctly
chosen number of linguistic terms for each specific variable at the synthesis
of the optimal rule base has a more significant impact on the efficiency of a
fuzzy system than the total number of rules.
To compare the effectiveness of the developed fuzzy controllers using the
first and the second modifications of the proposed information technology,
Figure 9.9 shows the flight trajectories over the mountainous terrain zT (x)
(absolute altitude values) of the QUAV with a control system based on the
given FCs. In turn, the detailed QUAV’s flight trajectories are presented in
Figure 9.10 (z = –2. . . 12 m; x = 0. . . 100 m) and Figure 9.11. (z = 5. . . 35 m; x
= 780. . . 900 m).
 9.4 Design and Structural Optimization of the Fuzzy Control System 225

Figure 9.9 QUAV’s flight trajectories over the mountainous terrain

Also, Figure 9.12 presents the transients of the flight control system of the
quadcopter (changing of the real values of flight altitude relative to the terrain
surface) with developed fuzzy controllers at flying over the mountainous
terrain with complex relief zT (x) (Figure 9.6). In turn, the detailed transients
of the flight control system are presented in Figure 9.13 (t = 0. . . 40 s) and
Figure 9.14 (zR = 3.5. . . 6.5 m; t = 100. . . 220 s). The simulation results
presented in Figures 9.9–9.14 are obtained for the QUAV with a flight altitude
control system based on the FCs: 1 – developed by the first modification;
2 – developed by the second modification. The flight altitude set value
zS (relative to the terrain surface) is 5 m, the QUAV speed ẋs along the
longitudinal-horizontal axis x over the mountainous terrain equals 4.3 m/s.
As Figure 9.9–9.14 shows, the QUAV’s flight control system based on
the FC developed using the first modification has higher accuracy and qual-
ity indicators of control compared to the same system based on the fuzzy
controller designed by the second modification. Moreover, the FC developed
using the first modification has a more straightforward structure, fewer rules
in the RB, and less computational costs were spent during its development,
which in general confirms the high effectiveness of this modification of the
proposed information technology.
226 Fuzzy Systems Design: Optimal Selection

Figure 9.10 Detailed QUAV’s flight trajectories (z = –2. . . 12 m; x = 0. . . 100 m)

Figure 9.11 Detailed QUAV’s flight trajectories (z = 5. . . 35 m; x = 780. . . 900 m)

 9.4 Design and Structural Optimization of the Fuzzy Control System 227

Figure 9.12 Transients of the altitude control system of the QUAV with developed FCs

Figure 9.13 Detailed transients of the altitude control system (t = 0. . . 40 s)

228 Fuzzy Systems Design: Optimal Selection

Figure 9.14 Detailed transients of the altitude control system (zR = 3.5. . . 6.5 m;
t = 100. . . 220 s)

In turn, if there is a need for additional improvement of the developed

fuzzy controllers after their synthesis and optimization of vector S of the LTs
number, it is possible to perform further structural and parametric optimiza-
tion. In particular, it is advisable to perform rule base reduction and optimize
the membership functions types and parameters [30, 39, 97].

9.5 Conclusion
The authors presented the development and study of the advanced infor-
mation technology for design and structural optimization of fuzzy systems
with optimal selection of linguistic terms number in this chapter. The appli-
cation of the developed information technology increases the FS accuracy
and performance, reduces the computational costs spent on the rule base
generation and optimization of parameters, and simplifies its software and
hardware realization. In particular, in this chapter, two modifications of the
given information technology for fuzzy systems development and structural
optimization with optimal selection of LTs number are developed and studied:
1) based on the sequential search approach; 2) based on the random search
approach. In turn, each modification uses a bioinspired ACO-based method
 References 229

and sequential search technique for automatic rule base design for each new
obtained variant of the vector S that defines numbers of linguistic terms. The
first modification uses the concept of sequential selection of the best values
of the number of LTs for the FS, starting from the first input x1 and ending
with the last output ym . The second modification uses the concept of creating
H random variants of vectors of the LTs number and choosing the best from
them.
The effectiveness study of both modifications of the developed informa-
tion technology is performed on a specific example, namely when designing
the fuzzy control system for the quadrotor unmanned aerial vehicle. In
particular, the development and structural optimization with optimal selection
of LTs number of the fuzzy controller for QUAV’s altitude control system is
carried out alternately using first and second modifications of the proposed
information technology. The conducted studies confirm, in general, the high
effectiveness of the presented information technology since its application
made it possible to develop a highly efficient QUAV’s fuzzy control system
with a fairly simple configuration and RB. The high efficiency is defined by
the obtained flight trajectories of the QUAV over the mountainous terrain with
complicated relief and control transients and low computational costs spent
at developing the studied fuzzy control system.
Moreover, when comparing the first and the second modifications with
each other, it is evident that the first modification has a higher efficiency than
the second since it allows to find the optimal configurations of fuzzy systems
(by the number of LTs for input and output variables) using lower computa-
tional costs. Also, we can conclude that when designing fuzzy systems, for
which simplicity of configuration, the small size of the RB, and simplicity
of software and hardware implementation are more important, it is more
advisable to use the first modification of this information technology. Since it
allows developing enough effective fuzzy systems with the simplest structure
and uses low computational costs at the same time. In turn, the second
modification can be quite effectively used to design FSs, for which high
accuracy and efficiency of solving the assigned problem (control or decision-
making) are paramount. Also, the configuration complexity, the RB size, and
the complexity of these systems’ software and hardware implementation are
of less importance.
Further research should be conducted towards developing and studying a
new specific modification of the proposed information technology for design
and structural optimization with optimal selection of LTs number for fuzzy
decision-making and control systems of the Takagi–Sugeno type.
230 Fuzzy Systems Design: Optimal Selection

References
[1] K. Knight, C. Zhang, G. Holmes, M.-L. Zhang (Eds.), ‘Artificial Intel-
ligence,’ Second CCF International Conference, ICAI 2019, Xuzhou,
China, August 22–23, 2019, Proceedings, Springer Singapore, 2019.
DOI 10.1007/978-981-32-9298-7.
[2] S. N. Vassilyev, A. Yu. Kelina, Y. I. Kudinov, F. F. Pashchenko, ‘Intel-
ligent Control Systems,’ in Procedia Computer Science, Vol. 103, pp.
623–628, 2017.
[3] C.C. Aggarwal, ‘Artificial Intelligence,’ Springer International Publish-
ing, 2021. DOI 10.1007/978-3-030-72357-6.
[4] G. Krivulya, I. Skarga-Bandurova, Z. Tatarchenko, O. Seredina, M.
Shcherbakova, E. Shcherbakov, ‘An intelligent functional diagnostics
of wireless sensor network,’ in Proceedings of 2019 International Con-
ference on Future Internet of Things and Cloud Workshops, FiCloudW
2019, pp. 135–139, 2019.
[5] L. A. Zadeh, ‘Fuzzy Sets,’ in Information & Control, 8, pp. 338–353,
1965.
[6] L. A. Zadeh, ‘The role of fuzzy logic in modeling, identification and
control,’ in Modeling Identification and Control, 15 (3), pp. 191–203,
1994.
[7] L. A. Zadeh, A. M. Abbasov, R. R. Yager, S. N. Shahbazova, M. Z.
Reformat, Eds., ‘Recent Developments and New Directions in Soft
Computing,’ STUDFUZ 317, Cham: Springer, 2014.
[8] J. Kacprzyk, ‘Multistage Fuzzy Control: A Prescriptive Approach,’ John
Wiley & Sons, Inc., New York, NY, USA, 1997.
[9] B. Kosko, ‘Fuzzy Systems as Universal Approximators,’ in IEEE Trans.
on Computers, Vol. 43, Iss. 11, pp. 1329–1333, 1994.
[10] Y. P. Kondratenko, O. V. Korobko, O. V. Kozlov, ‘Synthesis and Opti-
mization of Fuzzy Controller for Thermoacoustic Plant,’ in Recent
Developments and New Direction in Soft-Computing Foundations and
Applications, Studies in Fuzziness and Soft Computing 342, Lotfi A.
Zadeh et al. (Eds.), Berlin, Heidelberg: Springer-Verlag, pp. 453–467,
2016.
[11] J. M. Mendel, ‘Uncertain Rule-Based Fuzzy Systems, Introduction and
New Directions,’ Second Edition, Springer International Publishing,
2017.
[12] M. Solesvik, Y. Kondratenko, G. Kondratenko, I. Sidenko, V.
Kharchenko, A. Boyarchuk, ‘Fuzzy decision support systems in marine
 References 231

practice,’ in Fuzzy Systems (FUZZ-IEEE), 2017 IEEE International

Conference on Fuzzy Systems, pp. 1–6, 2017.
[13] Y. P. Kondratenko, O. V. Kozlov, ‘Mathematical Model of Ecopy-
rogenesis Reactor with Fuzzy Parametrical Identification,’ in Recent
Developments and New Direction in Soft-Computing Foundations and
Applications, Studies in Fuzziness and Soft Computing 342, Lotfi A.
Zadeh et al. (Eds.). Berlin, Heidelberg: Springer-Verlag, pp. 439–451,
2016.
[14] T. Bora, P. Chatterjee, S. Ghosh, ‘Fuzzy Logic Based Control Of Vari-
able Wind Energy System,’ in 2020 5th IEEE International Conference
on Recent Advances and Innovations in Engineering (ICRAIE), Jaipur,
India, 2020, pp. 1–5. DOI: 10.1109/ICRAIE51050.2020.9358376.
[15] Y. P. Kondratenko, O. V. Kozlov, ‘Mathematic Modeling of Reactor’s
Temperature Mode of Multiloop Pyrolysis Plant,’ in a book: Lecture
Notes in Business Information Processing: Modeling and Simulation in
Engineering, Economics and Management, K. J. Engemann, A. M. Gil-
Lafuente, J. M. Merigo, Eds., Vol. 115. Berlin, Heidelberg: Springer-
Verlag, pp. 178–187, 2012.
[16] Y.P. Kondratenko, O.V. Korobko, O.V. Kozlov, ‘Frequency Tuning
Algorithm for Loudspeaker Driven Thermoacoustic Refrigerator Opti-
mization,’ in K. J. Engemann, A. M. Gil-Lafuente, J. M. Merigo (Eds.),
Lecture Notes in Business Information Processing: Modeling and Sim-
ulation in Engineering, Economics and Management, Vol. 115, Berlin,
Heidelberg: Springer-Verlag, 2012, pp. 270–279. doi.org/10.1007/978-
3-642-30433-0_27.
[17] E. H. Mamdani, ‘Application of fuzzy algorithms for control of simple
dynamic plant,’ in Proceedings of IEEE, Vol. 121, pp. 1585–1588, 1974.
[18] E. H. Mamdani, S. Assilian, ‘An experiment in linguistic synthesis
with a fuzzy logic controller,’ in International journal of man-machine
studies, No. 7(1), pp. 1–13, 1975.
[19] A. Piegat, ‘Fuzzy Modeling and Control,’ New York, Heidelberg:
Physica-Verlag, 2001.
[20] R. Hampel, M. Wagenknecht, N. Chaker, ‘Fuzzy Control: Theory and
Practice, New York: Physika–Verlag,’ Heidelberg, 2000.
[21] Y. P. Kondratenko, O. V. Kozlov, O. V. Korobko, ‘Two Modifications
of the Automatic Rule Base Synthesis for Fuzzy Control and Decision
Making Systems,’ in J. Medina et al. (Eds), Information Processing and
Management of Uncertainty in Knowledge-Based Systems: Theory and
Foundations, 17th International Conference, IPMU 2018, Cadiz, Spain,
232 Fuzzy Systems Design: Optimal Selection

Proceedings, Part II, CCIS 854, Springer International Publishing AG,

pp. 570–582, 2018.
[22] Y. P. Kondratenko, O. V. Kozlov, G. V. Kondratenko, I. P. Atamanyuk,
‘Mathematical Model and Parametrical Identification of Ecopyrogenesis
Plant Based on Soft Computing Techniques,’ in Complex Systems:
Solutions and Challenges in Economics, Management and Engineering,
Christian Berger-Vachon, Anna María Gil Lafuente, Janusz Kacprzyk,
Yuriy Kondratenko, José M. Merigó, Carlo Francesco Morabito (Eds.),
Book Series: Studies in Systems, Decision and Control, Vol. 125, Berlin,
Heidelberg: Springer International Publishing, pp. 201–233, 2018.
[23] T. Takagi, M. Sugeno, ‘Fuzzy identification of systems and its applica-
tions to modeling and control,’ in IEEE Transactions on Systems, Man,
and Cybernetics, SMC–15, N 1, pp. 116–132, 1985.
[24] Y.P. Kondratenko, O.V. Kozlov, ‘Combined Fuzzy Controllers with
Embedded Model for Automation of Complex Industrial Plants,’ Shah-
naz N. Shahbazova, Janusz Kacprzyk, Valentina Emilia Balas, Vladik
Kreinovich (eds.) Recent Developments and the New Direction in Soft-
Computing Foundations and Applications. Studies in Fuzziness and Soft
Computing, Vol. 393, Springer, Cham, pp. 215–228, 2020.
[25] Z. Xiao, J. Guo, H. Zeng, P. Zhou, S. Wang, ‘Application of Fuzzy
Neural Network Controller in Hydropower Generator Unit,’ in J. Kyber-
netes, Vol. 38, No. 10, pp. 1709–1717, 2009.
[26] O. Kozlov, G. Kondratenko, Z. Gomolka, Y. Kondratenko, ‘Synthe-
sis and Optimization of Green Fuzzy Controllers for the Reactors of
the Specialized Pyrolysis Plants,’ in Kharchenko V., Kondratenko Y.,
Kacprzyk J. (eds) Green IT Engineering: Social, Business and Indus-
trial Applications, Studies in Systems, Decision and Control, Vol. 171,
Springer, Cham, pp. 373–396, 2019.
[27] O. Kosheleva, V. Kreinovich, ‘Why Bellman-Zadeh approach to fuzzy
optimization,’ in Appl. Math. Sci. Vol. 12, pp. 517–522, 2018.
[28] Y. P. Kondratenko, T. A. Altameem, E. Y. M. Al Zubi, ‘The optimiza-
tion of digital controllers for fuzzy systems design,’ in Advances in
Modelling and Analysis, AMSE Periodicals, Series A 47, pp. 19–29,
2010.
[29] Y. P. Kondratenko, A. V. Kozlov, ‘Generation of Rule Bases of Fuzzy
Systems Based on Modified Ant Colony Algorithms,’ in Journal of
Automation and Information Sciences, Vol. 51, Issue 3, New York:
Begel House Inc., pp. 4–25, 2019.
 References 233

[30] Y. P. Kondratenko, A. V. Kozlov, ‘Parametric optimization of fuzzy

control systems based on hybrid particle swarm algorithms with elite
strategy,’ in Journal of Automation and Information Sciences, Vol. 51,
Issue 12, New York: Begel House Inc., pp. 25–45, 2019.
[31] V.M. Kuntsevich et al. (Eds), ‘Control Systems: Theory and Appli-
cations,’ Book Series in Automation, Control and Robotics, River
Publishers, Gistrup, Delft, 2018.
[32] M. Jamshidi, V. Kreinovich, J. Kacprzyk, Eds., ‘Advance Trends in Soft
Computing,’ Cham: Springer-Verlag, 2013.
[33] Y. Kondratenko, G. Kondratenko, I. Sidenko, ‘Two-stage method of
fuzzy rule base correction for variable structure of input vector,’ in 2017
IEEE First Ukraine Conference on Electrical and Computer Engineering
(UKRCON), Kyiv, Ukraine, pp. 1043–1049, 2017.
[34] E. Ferreyra, H. Hagras, M. Kern, G. Owusu, ‘Depicting Decision-
Making: A Type-2 Fuzzy Logic Based Explainable Artificial Intel-
ligence System for Goal-Driven Simulation in the Workforce Allo-
cation Domain’, in 2019 IEEE International Conference on Fuzzy
Systems (FUZZ-IEEE), New Orleans, LA, USA, pp. 1–6, 2019. DOI:
10.1109/FUZZ-IEEE.2019.8858933.
[35] W. A. Lodwick, J. Kacprzhyk, Eds., ‘Fuzzy Optimization,’ STUDFUZ
254, Berlin, Heidelberg: Springer-Verlag, 2010.
[36] Y. Kondratenko, V. Korobko, O. Korobko, G. Kondratenko, O., Kozlov,
‘Green-IT Approach to Design and Optimization of Thermoacoustic
Waste Heat Utilization Plant Based on Soft Computing,’ Kharchenko
V., Kondratenko Y., Kacprzyk J. (Eds), Green IT Engineering: Com-
ponents, Networks and Systems Implementation. Studies in Systems,
Decision and Control, Vol. 105, Springer, Cham, pp. 287–311, 2017.
DOI: 10.1007/978-3-319-55595-9_14.
[37] K. Tanaka, H. O. Wang, ‘Fuzzy Control Systems Design and Analysis:
A Linear Matrix Inequality Approach,’ John Wiley & Sons, New York,
USA, 2001.
[38] D. Simon, ‘H∞ estimation for fuzzy membership function opti-
mization,’ in International Journal of Approximate Reasoning, 40,
pp. 224–242, 2005.
[39] Y. P. Kondratenko, D. Simon, ‘Structural and parametric optimization
of fuzzy control and decision making systems,’ in Zadeh L., Yager
R., Shahbazova S., Reformat M., Kreinovich V. (eds), Recent Devel-
opments and the New Direction in Soft-Computing Foundations and
234 Fuzzy Systems Design: Optimal Selection

Applications, Studies in Fuzziness and Soft Computing, Vol. 361,

Springer, Cham, pp. 273–289, 2018.
[40] Y. P. Kondratenko, E. Y. M. Al Zubi, ‘The Optimisation Approach
for Increasing Efficiency of Digital Fuzzy Controllers,’ in Annals of
DAAAM for 2009 & Proceeding of the 20th Int. DAAAM Symp.
“Intelligent Manufacturing and Automation,” Published by DAAAM
International, Vienna, Austria, pp. 1589–1591, 2009.
[41] B. Jayaram, ‘Rule reduction for efficient inferencing in similarity based
reasoning,’ in International Journal of Approximate Reasoning 48, no.
1, pp. 156–173, 2008.
[42] Y. Yam, P. Baranyi, C.- T. Yang, ‘Reduction of fuzzy rule base via
singular value decomposition,’ in IEEE Transactions on Fuzzy Systems,
7, no. 2, pp. 120–132, 1999.
[43] Y. P. Kondratenko, L. P. Klymenko, E. Y. M. Al Zu’bi, ‘Structural
Optimization of Fuzzy Systems’ Rules Base and Aggregation Models,’
in Kybernetes, Vol. 42, Iss. 5, pp. 831–843, 2013.
[44] D. Simon, ‘Evolutionary Optimization Algorithms: Biologically
Inspired and Population-Based Approaches to Computer Intelligence,’
John Wiley & Sons, 2013.
[45] O.V. Kozlov, Y.P. Kondratenko, ‘Bio-Inspired Algorithms for Optimiza-
tion of Fuzzy Control Systems: Comparative Analysis,’ in Yuriy P.
Kondratenko, Vsevolod M. Kuntsevich, Arkadiy A. Chikrii, Vyacheslav
F. Gubarev (Eds.) “Advanced Control Systems: Theory and Applica-
tions.” Series in Automation, Control and Robotics, River Publishers,
Gistrup, Denmark, pp. 83–128, 2021.
[46] W. Pedrycz, K. Li, M. Reformat, ‘Evolutionary reduction of fuzzy rule-
based models,’ in Fifty Years of Fuzzy Logic and its Applications,
STUDFUZ 326, Cham: Springer, pp. 459–481, 2015.
[47] M. Dorigo, M. Birattari, ‘Ant Colony Optimization,’ in Encyclopedia
of Machine Learning, Sammut C., Webb G.I. (eds.), Springer, Boston,
MA, 2011.
[48] B. Benhala, A. Ahaitouf, M. Fakhfakh, A. Mechaqrane, ‘New Adap-
tation of the ACO Algorithm for the Analog Circuits Design Optimiza-
tion’ in International Journal of Computer Science (IJCSI), Vol. 9, no. 3,
pp. 360–367, 2012.
[49] R. Gan, Q. Guo, H. Chang, Y. Yi, ‘Improved Ant Colony Optimization
Algorithm for the Traveling Salesman Problems,’ in Journal of Systems
Engineering and Electronics, pp. 329–333, 2010.
 References 235

[50] A. Nabi, N.A. Singh, ‘Particle swarm optimization of fuzzy logic

controller for voltage sag improvement,’ in Proceedings of 2016 3rd
International Conference on Advanced Computing and Communication
Systems (ICACCS), Vol. 01, pp. 1–5, 2016.
[51] S. Vaneshani, H. Jazayeri-Rad, ‘Optimized Fuzzy Control by Particle
Swarm Optimization Technique for Control of CSTR,’ in Interna-
tional Journal of Electrical and Computer Engineering, Vol. 5, No:11,
pp. 1243–1248, 2011.
[52] A. Engelbrecht, ‘A study of particle swarm optimization particle trajec-
tories,’ in Information Sciences, No 176(8), pp. 937–971, 2006.
[53] D. Simon, ‘Biogeography-Based Optimization,’ in IEEE Transactions
on Evolutionary Computation, 12(6), pp. 702–713, 2008.
[54] G. Thomas, P. Lozovyy, D. Simon, ‘Fuzzy Robot Controller Tuning with
Biogeography-Based Optimization,’ in Modern Approaches in Applied
Intelligence: 24th International Conference on Industrial Engineering
and Other Applications of Applied Intelligent Systems, IEA/AIE 2011,
Syracuse, NY, USA, June 28 – July 1, Proceedings, Part II, pp. 319–327,
2011.
[55] M. Huang, S. Shi, X. Liang, X. Jiao, Y. Fu, ‘An Improved Biogeography-
Based Optimization Algorithm for Flow Shop Scheduling Problem,’
in 2020 IEEE 8th International Conference on Computer Science and
Network Technology (ICCSNT), Dalian, China, pp. 59–63, 2020. DOI:
10.1109/ICCSNT50940.2020.9305008.
[56] S. Mirjalili, S. M. Mirjalili, A. Lewis, ‘Grey wolf optimizer.’ Adv. Eng.
Software, Vol., 69, pp. 46–61, 2014.
[57] B.P. Sahoo, S. Panda, ‘Improved grey wolf optimization technique for
fuzzy aided PID controller design for power system frequency con-
trol,’ J. Sustainable Energy, Grids and Networks, Vol. 16, pp. 278–299,
2018.
[58] T. Jayabarathi, T. Raghunathan, B.R. Adarsh, P.N. Suganthan, ‘Eco-
nomic dispatch using hybrid grey wolf optimizer,’ Energy, Vol. 111,
pp. 630–641, 2016.
[59] S. Khan, et al., ‘Design and Implementation of an Optimal Fuzzy Logic
Controller Using Genetic Algorithm,’ in Journal of Computer Science,
Vol. 4, No. 10, pp. 799–806, 2008.
[60] S. K. Oh, W. Pedrycz, ‘The Design of Hybrid Fuzzy Controllers Based
on Genetic Algorithms and Estimation Techniques,’ in J. Kybernetes,
Vol. 31, No. 6, pp. 909–917, 2002.
236 Fuzzy Systems Design: Optimal Selection

[61] R. Alcalá, J. Alcalá-Fdez, M. J. Gacto, F. Herrera, ‘Rule base reduction

and genetic tuning of fuzzy systems based on the linguistic 3-tuples
representation’, in Soft Computing 11, no. 5, pp. 401–419, 2007.
[62] F. L. Minku, T. Ludermir, ‘Evolutionary strategies and genetic algo-
rithms for dynamic parameter optimization of evolving fuzzy neural
networks,’ in Evolutionary Computation, The 2005 IEEE Congress,
Vol. 3, pp. 1951–1958, 2005.
[63] R. T. Alves, M. R. Delgado, H. S. Lopes, A. A. Freitas, ‘An Artificial
Immune System for Fuzzy-Rule Induction in Data Mining,’ in Yao X.
et al. (eds) Parallel Problem Solving from Nature - PPSN VIII. PPSN
2004. Lecture Notes in Computer Science, Vol 3242. Springer, Berlin,
Heidelberg, pp. 1011–1020, 2004.
[64] J. Zhu, F. Lauri, A. Koukam, V. Hilaire, ‘Fuzzy Logic Control Opti-
mized by Artificial Immune System for Building Thermal Condition,’
in Siarry P., Idoumghar L., Lepagnot J. (eds) Swarm Intelligence
Based Optimization, ICSIBO 2014, Lecture Notes in Computer Science,
Vol. 8472. Springer, Cham, pp. 42–49, 2014.
[65] A. Prakash, S. G. Deshmukh, ‘A multi-criteria customer allocation
problem in supply chain environment: An artificial immune system
with fuzzy logic controller based approach,’ in Expert Systems with
Applications, Volume 38, Issue 4, pp. 3199–3208, 2011.
[66] R. Menon, S. Menon, D. Srinivasan, L. Jain, ‘Fuzzy logic decision-
making in multi-agent systems for smart grids,’ in Computational Intel-
ligence Applications in Smart Grid (CIASG), 2013 IEEE Symposium,
pp. 44–50, 2013.
[67] O. Castillo, P. Ochoa, J. Soria, ‘Differential Evolution with Fuzzy
Logic for Dynamic Adaptation of Parameters in Mathematical Func-
tion Optimization.’ in Angelov P., Sotirov S. (eds) Imprecision and
Uncertainty in Information Representation and Processing. Studies in
Fuzziness and Soft Computing, Vol. 332. Springer, Cham, pp. 361–374,
2016.
[68] N. Quijano, K. M. Passino, ‘Honey bee social foraging algorithms for
resource allocation: theory and application,’ in Columbus: Publishing
house of the Ohio State University, 2007.
[69] D. H. Kim, C. H. Cho, ‘Bacterial foraging based neural network fuzzy
learning,’ Proceedings of the 2nd Indian International Conference on
Artificial Intelligence (IICAI – 2005), Pune: IICAI, pp. 2030–2036,
2005.
 References 237

[70] A. Melendez, O. Castillo, ‘Evolutionary optimization of the fuzzy inte-

grator in a navigation system for a mobile robot,’ in Recent Advances
on Hybrid Intelligent Systems, pp. 21–31, 2013.
[71] H. Ishibuchi, T. Yamamoto, ‘Fuzzy rule selection by multi-objective
genetic local search algorithms and rule evaluation measures in data
mining,’ in Fuzzy Sets and Systems 141, no. 1, pp. 59–88, 2004.
[72] G. Narvydas, R. Simutis, V. Raudonis, ‘Autonomous Mobile Robot
Control Using Fuzzy Logic and Genetic Algorithm,’ in IEEE Interna-
tional Workshop on Intelligent Data Acquisition and Advanced Com-
puting Systems: Technology and Applications, Dortmund, Germany,
pp. 460–464, 2007.
[73] O. Kozlov, ‘Optimal Selection of Membership Functions Types for
Fuzzy Control and Decision Making Systems,’ in Proceedings of the
2nd International Workshop on Intelligent Information Technologies
& Systems of Information Security with CEUR-WS, Khmelnytskyi,
Ukraine, IntelITSIS 2021, CEUR-WS, Vol-2853, pp. 238–247, 2021.
[74] O. Cordon, F. Gomide, F. Herrera, F. Hoffmann, L. Magdalena, ‘Ten
Years of Genetic Fuzzy Systems: Current Framework and New trends,’
in Fuzzy Sets and Systems, Vol. 141, Iss. 1, pp. 5–31, 2004.
[75] L. T. Koczy, K. Hirota, ‘Size reduction by interpolation in fuzzy rule
bases,’ in IEEE Transactions on Systems, Man, and Cybernetics, Part B:
Cybernetics, 27, no. 1, pp. 14–25, 1997.
[76] D. Driankov, H. Hellendoorn, M. Reinfrank, ‘An introduction to fuzzy
control,’ Springer Science & Business Media, 2013.
[77] J. P. Fernández, M. A. Vargas, J. M. V. García, J. A. C. Carrillo, J. J.
C. Aguilar, ‘Coevolutionary Optimization of a Fuzzy Logic Controller
for Antilock Braking Systems Under Changing Road Conditions,’ IEEE
Transactions on Vehicular Technology, 70, 2, pp. 1255–1268, 2021.
DOI: 10.1109/TVT.2021.3055142.
[78] Q. Suna, R. Li, P. Zhang, ‘Stable and Optimal Adaptive Fuzzy Control
of Complex Systems Using Fuzzy Dynamic Model,’ in J. Fuzzy Sets
and Systems, Vol. 133, pp. 1–17, 2003.
[79] Y.P. Kondratenko, O.V. Kozlov, L.P. Klymenko, G.V. Kondratenko,
‘Synthesis and Research of Neuro-Fuzzy Model of Ecopyrogenesis
Multi-circuit Circulatory System,’ Advance Trends in Soft Comput-
ing, M. Jamshidi, V. Kreinovich, J. Kazprzyk (Eds.), Series: Studies
in Fuzziness and Soft Computing, Vol. 312, pp. 1–14, 2014. DOI:
10.1007/978-3-319-03674-8_1.
238 Fuzzy Systems Design: Optimal Selection

[80] C. Von Altrock, ‘Applying fuzzy logic to business and finance,’ in

Optimus, 2, pp. 38–39, 2002.
[81] M. Pasieka, N. Grzesik, K. Kuźma, ‘Simulation modeling of fuzzy logic
controller for aircraft engines,’ in International Journal of Computing,
16(1), pp. 27–33, 2017.
[82] A. P. Rotshtein, H. B. Rakytyanska, ‘Fuzzy evidence in identification,
forecasting and diagnosis,’ Vol. 275, Heidelberg: Springer, 2012.
[83] D. Simon, ‘Sum Normal Optimization of Fuzzy Membership
Functions,’ in International Journal of Uncertainty: Fuzziness and
Knowledge-Based Systems, 10, pp. 363–384, 2002.
[84] A.M. Nazarenko, M.V. Karpusha, ‘Modeling and Identification in the
Problems of Multicriteria Optimization with Linear and Quadratic Per-
formance Criteria under Statistical Uncertainty,’ Journal of Automation
and Information Sciences, Volume 46, Issue 3, pp. 17–29, 2014. DOI:
10.1615/JAutomatInfScien.v46.i3.30
[85] Y. Kondratenko, P. Khalaf, H. Richter, D. Simon, ‘Fuzzy Real-Time
Multiobjective Optimization of a Prosthesis Test Robot Control System,’
in Yuriy P. Kondratenko, Arkadii A. Chikrii, Vyacheslav F. Gubarev,
Janusz Kacprzyk (Eds) Advanced Control Techniques in Complex
Engineering Systems: Theory and Applications, Dedicated to Profes-
sor Vsevolod M. Kuntsevich, Studies in Systems, Decision and Con-
trol, Vol. 203. Cham: Springer Nature Switzerland AG, pp. 165–185,
2019.
[86] J. Zhao, L. Han, L. Wang, Z. Yu, ‘The fuzzy PID control optimized
by genetic algorithm for trajectory tracking of robot arm,’ in 2016
12th World Congress on Intelligent Control and Automation (WCICA),
Guilin, China, pp. 556–559, 2016.
[87] Y.P. Kondratenko, O.V. Korobko, O.V. Kozlov, ‘PLC-Based Systems
for Data Acquisition and Supervisory Control of Environment-Friendly
Energy-Saving Technologies,’ chapter in a book: “Green IT Engineer-
ing: Concepts, Models, Complex Systems Architectures, Studies in
Systems, Decision and Control,” Vyacheslav Kharchenko, Yuriy Kon-
dratenko, Janusz Kacprzyk (Eds.), Book Series: Studies in Systems,
Decision and Control, Vol. 74, Berlin. Heidelberg: Springer Interna-
tional Publishing, pp. 247–267, 2017.
[88] Y. Khaluf, S. Gullipalli, ‘An Efficient Ant Colony System for Edge
Detection in Image Processing,’ in Proceedings of the European Con-
ference on Artificial Life, pp. 398–405, 2015.
 References 239

[89] R.-M. Chen, Y.-M. Shen, C.-T. Wang, ‘Ant Colony Optimization
Inspired Swarm Optimization for Grid Task Scheduling,’ in 2016 Inter-
national Symposium on Computer, Consumer and Control (IS3C),
pp. 461–464, 2016.
[90] V.L. Timchenko, D.O. Lebedev, ‘Optimization of Processes of Robust
Control of Quadcopter for Monitoring of Sea Waters,’ Journal of
Automation and Information Sciences, NY., Begell House, Vol. 51, Issue
2, pp. 1–10, 2019.
[91] N. Ben, S. Bouallègue, J. Haggège, ‘Fuzzy gains-scheduling of an inte-
gral sliding mode controller for a quadrotor unmanned aerial vehicle,’
Int. J. Adv. Comput. Sci. Appl., Vol. 9, no. 3, pp. 132–141, 2018.
[92] Y. Kondratenko, ‘Robotics, Automation and information systems:
Future perspectives and correlation with culture, Sport and life science,’
Lecture Notes in Economics and Mathematical Systems, Vol. 675,
Springer Verlag, pp. 43–55, 2014. DOI: 10.1007/978-3-319-03907-7_6.
[93] A. Eltayeb, M. F. Rahmat, M. A. M. Basri, M. A. M. Eltoum, S. El-
Ferik, ‘An Improved Design of an Adaptive Sliding Mode Controller
for Chattering Attenuation and Trajectory Tracking of the Quadcopter
UAV,’ IEEE Access, Vol. 8, pp. 205968–205979, 2020.
[94] R. Duro et al. (Eds). ‘Advances in Intelligent Robotics and Collaborative
Automation.’ River Publishers, Gistrup, 2015. DOI: https://doi.org/10.1
3052/rp-9788793237049.
[95] Y.P. Kondratenko, et al (Eds). ‘Advanced Control Systems: Theory and
Applications.’ River Publishers, Gistrup, 2021.
[96] I.P. Atamanyuk, V.Y. Kondratenko, O.V. Kozlov, Y.P. Kondratenko,
‘The Algorithm of Optimal Polynomial Extrapolation of Random Pro-
cesses,’ Lecture Notes in Business Information Processing: Modeling
and Simulation in Engineering, Economics and Management., K. J.
Engemann, A. M. Gil-Lafuente, J. M. Merigo (Eds.), Berlin, Heidelberg:
Springer-Verlag, Vol. 115, pp. 78–87, 2012.
[97] D. Simon, ‘Design and rule base reduction of a fuzzy filter for the
estimation of motor currents,’ in International Journal of Approximate
Reasoning, 25, pp. 145–167, 2000.
 10
Analysis of the Dynamics
and Controllability of an Autonomous
Mobile Robot with a Manipulator

Ashhepkova Natalja1 , Zbrutsky Alexander2 ,

and Koshevoy Nicolay3
1 Department of Mechanotronics, Oles Honchar Dnipro National University,
Ave. Gagarin, 72, Dnipro, Ukraine, 49010
2 Department of Aircraft Control Systems, National Technical University

of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”, Str. Botkin, 1, Kyiv,

Ukraine, 03056
3 Department of Intelligent Visual Systems and Engineering of Quality,

National Aerospace University “Kharkov Aviation Institute”, st. Chkalov,

17, Kharkov, Ukraine, 61070
E-mail: [email protected]; [email protected];
[email protected]

Abstract
The results of the study of the dynamics and control of an autonomous mobile
robot with a manipulator are presented. An all-wheel drive model with a
four-wheeled chassis layout and equipped with a four-link manipulator is
considered. The design of the manipulator consists of a docking ring rotating
around a vertical axis and three rod links of the arm, connected by rotary
kinematic pairs of the fifth class. The mathematical model of controlled
movement of autonomous mobile robot with a manipulator is proposed. With
relative motion of the manipulator, the tensor of inertia of an autonomous
mobile robot in the coordinate system associated with the chassis is non-
diagonality and non-stationarity. Simplifications of the mathematical model
for several driving modes are substantiated. An algorithm for integrating the
equations of dynamics is developed, taking into account certain features.

241
242 Analysis of the Dynamics and Controllability of an Autonomous Mobile Robot

Based on the results of mathematical modeling, an algorithm for adaptive

control of autonomous mobile robot with a manipulator is proposed.
Keywords: autonomous mobile robot, non-diagonal tensor of inertia, dynam-
ics, adaptive control.

10.1 Introduction
The development of new technologies makes it possible to use manipulators
(M) for work in extreme conditions for humans. For example, the use of M
and autonomous mobile robots (AMR) for eliminating the consequences of
environmental and man-made disasters, the operation of remotely controlled
M for working with radioactive substances, the use of M for assembling
structures in space or on the seabed, etc. For work in extreme conditions, as
a rule, remote-controlled robots and M. are used. In many cases, a remote-
controlled arm-manipulator is installed on a controlled mobile platform.
This expands the service characteristics of M, but imposes more stringent
requirements on the design of control and monitoring systems. An example
of such structures is: a manipulator arm is placed on a spacecraft, on a ship,
on a wheeled chassis, etc. In all of the above cases, the movement of the
base affects the kinematics and dynamics of the manipulator arm, increasing
the positioning error and deviation of the gripper pole from the specified
trajectory. Dynamic processes for M on a mobile platform can be divided
into three modes of movement:
• controlled movement of the platform along a given route with “check-
in” M;
• performing technological operations M with the platform stopped;
• M performs technological operations with a controlled movement of the
platform.
The controlled movement of a platform with a “locked” manipulator has been
well studied for various chassis designs. In [1], the dynamics of an all-wheel
drive 4-wheeled bogie when turning is investigated. The synthesis of the law
of control of the plane motion of a transport robot is a typical task: several
methods of stabilization of motion along a given trajectory are known. In [2],
the authors propose a modification of the system of dynamic equations with
the change of coordinates and the transformation of outputs to chain forms.
In [3], a transformation of the system of equations to the Cauchy normal form
was applied with subsequent differentiation with respect to a new independent
variable [3–5]. In [2, 6], the equations of the bogie dynamics are relative to
road coordinates, which well fix deviations from a given trajectory.
 10.1 Introduction 243

When performing technological operations M, which is installed on a

stopped platform, it is advisable to consider the equation of dynamics with
respect to a fixed coordinate system associated with the docking point M
and the platform. In this case, the solution of the problems of kinematic and
dynamic analysis of M depends on the ratio of the inertial characteristics of
the platform and M with a load. The load can be a manipulated object, probe,
drill, bucket, etc. In [7–10], two approaches to formulating the equations
of the dynamics of M are considered: the Newton–Euler method and the
Lagrange–Euler method. To simulate the processes of dynamics and control
of M, it is advisable to use the mathematical packages Mathcad [11–13] or
Matlab [14–16]. For the effective execution of M technological operations in
extreme conditions, it is necessary at the design stage to automatically carry
out kinematic and dynamic analysis [11], determine the boundaries of the
working space [12], analyze the optimality of configurations at the points
of the working space, and ensure the positioning accuracy of the gripper
[13]. In [14], pseudo symbolic dynamic modeling (PSDM) is presented for
creating simplified dynamic models M, the design of which contains up to
7 degrees of mobility. The presented algorithm allows to generate code in
real time, simulate dynamics, and improve the efficiency of the model by
eliminating the least important elements. In addition, the authors of [14] have
developed an implementation of the algorithm in the MATLAB environment,
which is publicly available. The authors of [15] developed a virtual model
of kinematics and dynamics M in MATLAB & SIMULINK. The presented
model uses a PID controller, and the dynamics equations are obtained by the
Lagrange–Euler method. The work [16] is devoted to the dynamic modeling
of three-link spatial M using symbolic and numerical methods. An algorithm
based on the Newton–Euler method is proposed to formulate the equations of
dynamics in the form of a state space. The algorithm is implemented in the
Maple system, the modeling is carried out in MATLAB & SIMULINK.
When developing the design of the AMR arm-manipulator, it is advisable
to investigate the efficiency of technological operations with a movable
platform. Let’s analyze the methods for studying the dynamics and control
of a manipulator arm on a mobile platform, for example, on a ship, spacecraft
or ground transport. In [17–19], a multi-mass system is considered, which
consists of a spacecraft and M. The manipulator arm mounted on the ship’s
deck was considered in [20–23]. It is generally believed that the M dynamics
affects the platform, but does not cause it to move. However, the results
presented in [17] prove that the controlled motions of M cause oscillations of
the spacecraft and an increase of 43% in energy consumption for its angular
244 Analysis of the Dynamics and Controllability of an Autonomous Mobile Robot

stabilization. In [24, 25], the dynamic equations of a manipulator arm on a

movable platform are derived. It is believed that the inertia of the platform
is large enough and is not affected by the relative motion of the manipulator.
In [26], the analysis of the inertia tensor AMR with M with different con-
figurations of their mutual position showed that the values of the centrifugal
moments of inertia of the system with relative motion M are comparable to
the value of the axial moments of inertia of the system, even if the mass M
with a load is less than 20% of the platform mass. So, the relative movement
of the manipulator induces linear or angular movements of the platform.
Therefore, in the general case, the tensor of inertia of such a system must
be taken non-diagonality and non-stationarity, and the mathematical model
of the dynamics of AMR with M must take into account the cross-coupling
of control channels.
AMR with M, as a control object, is a multichannel, multiply connected,
essentially nonlinear dynamic system. The dynamics and control of robots
were studied by S. Burdakov, I. Miroshnik, E. Yurevich, Yu. Martynenko,
B. Topchiev and others. However, taking into account the nonlinearity and
complexity of the control system was carried out only for particular solutions.
For AMR with an adaptive control system, operating under not predetermined
environmental conditions, the driving system must be tested on a model.
When designing AMR, it is economically expedient to carry out not only
mathematical, but also simulation modeling.
Thus, the study of the dynamics of AMR with M by methods of mathe-
matical and simulation modeling is an urgent scientific and applied problem.

10.2 Results of Study of Dynamics and Control of

Autonomous Mobile Robots
10.2.1 Problem Statement
The expansion of the scope of application and the complication of tasks
for modern AMR determines the presence of structural elements that are
movable relative to the robot platform: a manipulator, a probe, a drill, a
bucket, etc. During the operation of the AMR, these elements move relative
to the platform, changing the geometry of the masses of the system. As an
example, AMR of variable configuration, let’s consider AMR with M. The
peculiarities of the dynamics of AMR with M, as a system of bodies of
variable configuration, are:
• change in the position of the center of mass of the system with relative
movement of the manipulator,
10.2 Results of Study of Dynamics and Control of Autonomous Mobile Robots 245

• commensurability of non-diagonality and diagonal elements of the ten-

sor of inertia, calculated with respect to the axes of the basic coordinate
system CXC YC Z C , associated with the center of mass of the AMR
platform.
The AMR operation cycle can be divided into several stages:
• movement of the AMR platform from the start point S to the finish point
F with a stationary manipulator in compliance with the requirements of
optimality (speed, productivity or economy, etc.);
• performing technological operations with a manipulator while the plat-
form is stopped;
• return of the AMR platform to the starting point S or movement to a
predetermined end point K with a stationary manipulator.
At the first stage of work, AMR is a cart with an adaptive control system that
moves in a predetermined environment. To ensure the stable AMR movement
from the start point S to the finish point F with a stationary manipulator in
compliance with the requirements of optimality, it is necessary to perform
mathematical and simulation modeling.
When returning the AMR platform, it is necessary to take into account
the change in the mass distribution of the structure relative to the center of
mass of the platform due to the presence of objects to be manipulated, soil
samples, water samples, etc. with a stationary manipulator.
Object of research: AMR with M, that is, a controlled wheeled cart, equipped
with a movable M and information devices (camera, microphone, dosimeter,
etc.). The control system allows to move in a limited space (working area)
along a given trajectory using GPS navigation or orientation on a map or
beacons. Within the limits of the working area, possible static and (or) moving
obstacles that must be overcome with the least deviation from the specified
route.
The aim of research is to increase the efficiency of using AMR with M, which
acts in not predetermined environmental conditions.
To achieve the aim, the research objectives were formulated:
• determination of the features of AMR dynamics with a manipulator;
• development of an algorithm for integrating the current state of the
equations of dynamics;
• assessment of the interconnection of control channels;
• development of an adaptive control algorithm.
246 Analysis of the Dynamics and Controllability of an Autonomous Mobile Robot

Research methods: conclusions are formulated in the work, scientific assump-

tions and recommendations are based on the fundamental provisions of
theoretical mechanics, the theory of differential equations and control
theory.

10.2.2 Design of an Autonomous Mobile Robot with a

Manipulator
A diagram of the AMR design with a manipulator is shown in Figure 10.1.
The design consists of an all-wheel drive 4-wheeled platform AMR, and an
anthropomorphic manipulator composed of a ring rotating around a vertical
axis and rod links of the arm, connected by rotational kinematic pairs of the
fifth class.

Figure 10.1 Diagram of the AMR design with a manipulator [26]

10.2 Results of Study of Dynamics and Control of Autonomous Mobile Robots 247

To analyze the results of mathematical modeling of the dynamics of AMR

with M, let’s assume the following:
• AMR platform—absolutely rigid body with uniform mass distribution
with density ρ=const; development of an algorithm for integrating the
current state of the equations of dynamics;
• the links of the manipulator are absolutely solid, rigid and have a
uniform mass distribution with a density ρ=const.
Further, for definiteness, when creating a mathematical model of AMR of
variable configuration, an all-wheel drive 4-wheeled platform AMR with a
manipulator is considered as an example.
Let’s introduce the following right coordinate systems (Figure 10.1):
AXY Z—inertial coordinate system.
CXC YC ZC —movable base coordinate system. The origin is connected to
the point C by the center of mass of the platform. The axes are parallel to the
main central axes of inertia of the AMR platform. Here CZC is perpendicular
to the platform movement plane, coincides with the local vertical and is
directed upward to the manipulator ring, CXC is located in the platform
movement plane and is directed towards the movement, the CYC axis is
located in the platform movement plane and complements the coordinate
system to the right.
OX0 Y0 Z0 —moving coordinate system. The origin is connected to point
O–by the center of mass of the system of bodies. The axes are parallel to the
axes of the CXC YC ZC coordinate system.
O1 X1 Y1 Z1 —moving coordinate system is connected. The origin is con-
nected to the point O1 –by the center of mass of the ring. The axes coincide
with the main central axes of inertia of the ring. In the initial position of the
manipulator ring, the axes of coordinate systems O1 X1 Y1 Z1 are parallel to
the axes of the CXC YC ZC coordinate system.
Oi Xi Yi Zi (for i = 2, 3, 4)—connected moving coordinate systems. The
origin is connected to the point Oi –by the center of the kinematic pair.
The Oi Xi axes coincide with the longitudinal axes of the rod links of the
manipulator arm, in the initial position of the manipulator links the axes of
the Oi Xi Yi Zi coordinate systems are parallel to the axes of the CXC YC ZC
coordinate system.
MXM YM ZM —movable base coordinate system. The origin is connected
to the point M—by the center of mass of the manipulator. In the initial
position of the manipulator, the axes of the MXM YM ZM coordinate systems
are parallel to the axes of the CXC YC ZC coordinate system.
248 Analysis of the Dynamics and Controllability of an Autonomous Mobile Robot

10.2.3 Dynamics Analysis

The mathematical model is compiled taking into account the non-diagonality
of the inertia tensor AMR with M. Assessment of the elements of the inertia
tensor AMR with a manipulator with different configurations of their mutual
position, if the mass of the manipulator with a load is from 10% to 30% of
the mass of the AMR platform, carried out in [26].
Let’s consider the controlled motion of the AMR with M taking into
account the non-diagonality of the tensor of inertia, if the main central axes
of inertia of the AMR change orientation relative to the vectors of the control
torques. The motion of AMR with M is characterized by the speed of motion
v̄˙ 0 along the trajectory of the pole O0 —the center of mass of the system and
by the vector of angular velocity Ω̄.
The mathematical model of the motion of a dynamic system, taking
into account the nonstationarity and non-diagonality of the tensor of inertia,
describes:
• movement of the center of mass of the AMR system with M;
• angular movement AMR with M;
• relative movement of the manipulator;
• equations of forces acting on AMR with M;
• equations of executive bodies of the control system.
1) Equations of motion of the center of mass of the AMR system with M
The equations of motion of the center of mass of the AMR system with
M when moving along the trajectory in the inertial coordinate system has the
form:
 2 ) * ' ˜2
˜0 Ω̄ d p̄
m · ddtr̄2c + ddt × p̄co1 + Ω̄ × Ω̄ × p̄co1 + m< · 1dto1M 2 +
' ) * ˜ ˜
+ 2 w̄1 + Ω̄ × d1 p̄dto1M + d1dtw̄1 × p̄o1M + Ω̄ × (w̄1 × p̄o1M ) +

+Ω̄ × (w̄
) 1 × p̄o1M )* + (Ω̄ × w̄1 ) × p̄o1M + 
+w̄1 × Ω̄ × p̄o1M +w̄1 × (w̄1 × p̄o1M )] = v F̄v .
(10.1)
For the sections of the trajectory on which the AMR moves with an
immovable manipulator, let’s write formula (10.1) in the form:
1 2
d2 r̄c d˜0 Ω̄ ) * !
m· 2
+ × p̄co1 + mM Ω̄ × Ω̄ × p̄CM = F̄v . (10.2)
dt dt v
If the AMR moves with a slow angular motion of the manipulator, from
formula (10.1) let’s obtain:
10.2 Results of Study of Dynamics and Control of Autonomous Mobile Robots 249
1 2 1
d2 r̄c d˜0 Ω̄ d˜21 p̄o1M ) * d˜1 p̄o1M
m· 2
+ × p̄ co1 + m M · 2
+ 2 w̄1 + Ω̄ × +
dt dt dt dt
2
d˜1 w̄1 !
+ × p̄o1M = F̄v . (10.3)
dt v

Equation (10.1) is the equation of motion of the center of mass of the

AMR system with a manipulator in the general case, and Equations (10.2)
and (10.3) are separate cases if the manipulator does not move or moves with
a limited speed.
2) The equation of AMR with M angular motion
The equations of angular motion of the AMR with M when moving along
the trajectory in the inertial coordinate system has the form:

d˜20 d˜0 Ω̄
p̄ oo1 · (mM p̄ o1M + m c p̄ o1c ) + (mM p̄o1M + mc p̄o1c ) +
dt2 dt
d˜0
+ 2Ω̄ × (mM p̄o1M + mc p̄o1c )
dt
8 9
+ Ω̄ × Ω̄ × (mM p̄o1M + mc p̄o1c )

! d˜20 p̄o1j d˜0 Ω̄ d˜0 p̄o1j
++ p̄o1j × mj + × p̄ o1j + 2 Ω̄ × + Ω̄
dt2 dt dt
j

) *
× Ω̄ × p̄o1j

! d˜20 p̄o1i d˜0 Ω̄ d˜0 p̄o1i
++ p̄o1i × mi + × p̄o1i + 2 Ω̄ × + Ω̄
dt2 dt dt
i

) *
× Ω̄ × p̄o1i = M̄o . (10.4)

3) Equation of manipulator relative motion

Let’s write the equation of the relative motion of the manipulator in the
inertial coordinate system:
250 Analysis of the Dynamics and Controllability of an Autonomous Mobile Robot

 d˜20 p̄oc d˜0 p̄oc d˜0 Ω̄

j p̄o1j × mj {v̄o + × × (p̄oc + p̄o1j ) +
˙ + 2 Ω̄ +
dt2 dt dt
8 9 d˜2 p̄o1j d˜ p̄ d˜ w̄
+Ω̄ × Ω̄ × (p̄oc + p̄o1j ) + 1 2 + Ω̄ × 1dto1j + 1dt 1 × p̄o1j +
dt

˜
d1 p̄o1j 8 9
+ 2w̄1 × + (Ω̄ × w̄1 ) × p̄o1j + w̄1 × (w̄1 + Ω̄) × p̄o1j
dt
= M̄01 .
(10.5)
4) Equation of executive bodies of the control system
Angular stabilization of the AMR manipulator, as a system of bodies of
variable configuration, occurs under the action of an active relay-type control
system. The main vector of the control torque, relative to point C: M̄cU =
M̄ · χ,
) *T
where M̄ = MxU MyU MzU – vector of moments of the control
motors, χ – control function:
⎧
⎪
⎪ 1, j ≤ −1, ddtj > 0; j ≤ −m, ddtj < 0;
⎪
⎨
χ= 0, −1 < j < m, ddtj > 0; −m < j < 1, ddtj < 0;
⎪
⎪
⎪
⎩
−1, j ≥ m, ddtj > 0; j ≥ 1, ddtj < 0.
⎛ (10.6)
⎞
Ωx
Vector of control actions j̄ = T · Ω̄ + ξ¯ , where Ω̄ = ⎝ Ωy ⎠,
Ωz
⎛ ⎞
ξx = ϕ
ξ¯ = ⎝ ξy = ψ ⎠.
ξz = υ
The equations for the on and off lines for each control channel are as
follows:
±gi = Ti · Ωi + ξi ;
±ni · gi = Ti · Ωi + ξi ;
i = x, y, z,
where0 ≤ ni = const ≤ 1 – the delay coefficient of the control system for
each channel; gi = const – restriction on motion parameters for each control
channel; Ti – time constants.
It was not possible to solve the system of equations of the mathematical
model by standard methods of numerical integration, because, in addition to
10.2 Results of Study of Dynamics and Control of Autonomous Mobile Robots 251

Figure 10.2 Relay static characteristic

the indicated features at the switching points jof f , jon , the control functions
χi have several values (Figure 10.2).
It is proposed to solve the system of equations by the Runge–Kutta–
Fehlberg method [27]. In this case, on the sections of motion corresponding
to χi =const, numerical integration is carried out with a certain constant step
h. If |j-jof f |≤E (or |j-jon |≤E , where E = const – the vicinity of the switching
point of the relay characteristic), then the integration step decreases and is
equal to h1 =h/C. The most common is the so-called double step method with
C=2n , n = const, n>0.
If the condition |j-jof f |≤E/C (or |j-jon |≤E/C) let’s assume that the switch-
ing of the relay characteristic takes place and χi changes the value in
accordance with the control law. The values of the constants h1 , h, E, C, n are
chosen based on the initial conditions. The structural diagram of the algorithm
for solving the system of equations by the Runge–Kutta–Fehlberg method is
shown in Figure 10.3. The use of the Runge–Kutta–Fehlberg method makes it
possible to create software for identifying the current dynamic state of AMR
with M.
The results of mathematical modeling of the AMR dynamics without
taking into account the non-diagonality and non-stationarity of the tensor of
inertia are given in Table. 10.1 (t – time; q1 (t) – the generalized coordinate of
the relative motion of the manipulator; αx (t), αy (t), αz (t) – the angular coor-
dinates of the system of bodies relative to the coordinate system CXC YC ZC ;
ω x (t), ω y (t), ω z (t) – the angular velocities of the system of bodies relative to
the coordinate system CXC YC ZC ; χx , χy , χz – control functions).
The results of mathematical modeling of the AMR dynamics taking into
account the non-diagonality and non-stationarity of the tensor of inertia are
252 Analysis of the Dynamics and Controllability of an Autonomous Mobile Robot

BEGIN

Input of initial data

Calculation of the elements

of the tensor of inertia

Defining control function

Yes Yes

No No
Integration with step ,

Integration with step ,

No No
Yes Yes No

Output of results Yes

Switching
, ,…, , , ,…, END

Figure 10.3 Algorithm for solving the problem of dynamics of AMR with M by the Runge–
Kutta–Fehlberg method

given in Table 10.2(t – time; q1 (t)–generalized coordinate of the relative

motion of the manipulator; αx (t), αy (t), αz (t) – angular coordinates of the
system of bodies relative to the coordinate system CXC YC ZC ; ω x (t), ω y (t),
ω z (t) – angular velocities of the system of bodies relative to the coordinate
CXC YC ZC ; χx , χy , χz – control functions).
The non-diagonality of the tensor of inertia of a system of bodies relative
to the base coordinate system determines:
• occurrence of inconsistency of the main central axes of inertia of the
system of bodies with the axes associated with the center of mass of the
AMR platform by the basic CXC YC ZC coordinate system,
• inconsistency of control actions with the directions of the main central
axes of inertia of the system,
• interdependence of control channels.
10.2 Results of Study of Dynamics and Control of Autonomous Mobile Robots 253

Table 10.1 Results of mathematical modeling of AMR dynamics without taking into account
the non-diagonality and non-stationarity of the tensor of inertia
t, c q1 , αx , ω x .10−3 , χx αy , ω y .10−3 , χy αz .10−3 , ω z .10−3 , χz
rad rad rad /s rad rad /s rad rad /s
0 0 0 0 0 0 0 0 15,71 17,45 −1
0,03 0 0 −0,1 1 0 0,12 −1 15,71 −17,45 1
18,03 0,05 −0,24 −0,1 1 −0,18 −0,1 1 −15,71 −17,45 1
18,05 0,05 −0,24 1,0 −1 −0,18 0,05 0 −15,71 17,45 −1
36,05 0,11 1,47 1,0 −1 −0,05 0,09 0 15,71 17,45 −1
36,08 0,11 1,47 −0,1 1 −0,05 −0,1 1 15,71 −17,45 1
54,08 0,16 1,24 −0,1 1 −0,01 0,02 0 −15,71 −17,45 1
54,11 0,16 1,37 1,0 −1 −0,01 0,14 −1 −15,71 17,45 −1
72,11 0,22 3,08 1,0 −1 0,23 0,18 −1 15,71 17,45 −1
72,14 0,22 3,08 −0,1 1 0,23 0,06 0 15,71 −17,45 1
90,14 0,27 2,85 −0,1 1 0,38 0,11 −1 −15,71 −17,45 1

Table 10.2 Results of mathematical modeling of AMR dynamics taking into account the
non-diagonality and non-stationarity of the tensor of inertia
t, c q1 , αx , ω x .10−3 , χx αy , ω y .10−3 , χy αz .10−3 , ω z .10−3 , χz
rad rad rad /s rad rad /s rad rad /s
0 0 0 0 0 0 0 0 15,71 17,45 −1
0,03 0 0 −0,1 1 0 −0,1 1 15,71 −17,45 1
18,03 0,05 −0,24 −0,1 1 −0,02 −0,1 1 −15,71 −17,45 1
18,05 0,05 −0,24 0,1 −1 −0,02 0,1 −1 −15,71 17,45 −1
36,05 0,11 0,14 0,1 −1 0,01 0,1 −1 15,71 17,45 −1
36,08 0,11 0,14 −0,1 1 0,01 −0,1 1 15,71 −17,45 1
54,08 0,16 −0,14 −0,1 1 −0,01 −0,1 1 −15,71 −17,45 1
54,11 0,16 −0,14 1,0 −1 −0,01 1,0 −1 −15,71 17,45 −1
72,11 0,22 0,13 1,0 −1 0,01 1,0 −1 15,71 17,45 −1
72,14 0,22 0,13 −0,1 1 0,01 −0,1 1 15,71 −17,45 1
90,14 0,27 −0,12 −0,1 1 −0,01 −0,1 1 −15,71 −17,45 1

10.2.4 Analysis of AMR with M Controllability

The results of mathematical modeling of linear motion of AMR are given
in [26] and prove that the growth of q1 (t) from 0 to 0.27 rad causes an
increase in angular coordinates and velocities in directions perpendicular to
254 Analysis of the Dynamics and Controllability of an Autonomous Mobile Robot

the direction of motion. The control system tries to work out the disturbances
and prevent the growth of angular velocities in the directions perpendicular
to the direction of motion. Despite the operation of the control system, due
to the non-diagonality and non-stationarity of the tensor of inertia and cross-
coupling of control channels, after 90 s of the AMR movement along the
trajectory, the angular coordinates reach the values αx = 2.85 rad, αy = 0.38
rad. Thus, AMR loses its vertical position, in the conditions of autonomous
operation it leads to the impossibility of performing technological
operations.
If the magnitude and direction of action of the control actions is deter-
mined taking into account the cross-connection of control channels due to the
non-diagonality and non-stationarity of the AMR inertia tensor relative to the
coordinate system CXC YC ZC , then the angular coordinates reach the values
αx = −0.12 rad, αy = −0.01 rad. In this case, the growth of q1 (t) from 0
to 0.27 rad also causes the occurrence of angular coordinates and velocities
in the directions perpendicular to the direction of motion, but AMR does not
lose its vertical position and continues to perform technological operations.
The effectiveness of manipulator control is largely determined by teach-
ing methods, applied methods and means of adaptation. The control of the
grab of the AMR manipulator is carried out by the operator remotely. The
design provides for three data transmission channels (fiber optic wire, Wi-fi
and protected radio communication). Correction of movements and control
of the results of the manipulator’s activity is carried out using a video
surveillance system.
The operating conditions of AMR with a manipulator are often not only
not known a priori, but can also vary unpredictably over a wide range. The
reasons for the uncertainty and nonstationarity of these conditions are: 1) the
lack of information about the properties of the external environment; 2) work
in technical limitations, natural spread and drift of parameters of sensory and
motor systems; 3) in the occurrence of obstacles and computational errors in
communication and control channels.
Usually an adaptive control system has to adapt to the conditions of the
external environment. That is why work with adaptive control and elements
of artificial intelligence is significantly superior in terms of the possibilities of
working with programmed control: they can adequately respond to changes
in the external environment, adapt to the drift of environmental parameters,
recognize and avoid obstacles, identify target objects, and determine their
characteristics.
10.2 Results of Study of Dynamics and Control of Autonomous Mobile Robots 255

The dynamic characteristics of the AMR system with a manipulator

change significantly during the operation cycle. Before generating a com-
mand to the actuators of the chassis and M, the control system has to
identify the current state of AMR with M. After which the command is
corrected and the distribution of influences of AMR with M on the actuators
takes into account the cross-connection of the control channels due to the
non-diagonality and non-stationarity of the inertia tensor.
In this regard, an algorithm for adaptive control of AMR with M is
proposed, which provides for the study and assessment in real time of the
dynamic characteristics of AMR (identification problems) and determination
of control parameters from the obtained model (synthesis problem).

Figure 10.4 Block diagram of the AMR adaptive control algorithm with M
256 Analysis of the Dynamics and Controllability of an Autonomous Mobile Robot

According to [28], the AMR model can be represented as an “input -

output” equation:
y(t) = fy [y(t − 1), y(t − 2), ..., u(t − 1), u(t − 2), ..Q(t), Θ(t)] ,
where y – output variable u – control, Θ – disturbances, fy – some function,
Q – parameters of AMR with manipulator, t=1,2,. . . n – discrete time.
The recurrent type adaptive control algorithm is written as
u(t) = fu [u(t − 1), y(t), Ue (t), Q(t)] ,
where Ue – desirable trajectory of AMR movements with manipulator, fu –
control function depending on the introduced law.
The adaptation criterion is to minimize the deviation of the AMR state
vector from the desired trajectory, which can be written in general form:
I = Φ(Ue (t) − y(t)) → min.
The block diagram of the proposed algorithm for adaptive control of
AMR with a manipulator is shown in Figure 10.4.
The proposed control algorithm will make it possible to equip work
with means of automatic learning (self-learning) and adaptation, which turn
their control system into an extremely flexible tool for organizing purposeful
behavior under conditions of uncertainty.

10.3 Conclusion
The peculiarities of the dynamics of AMR with M are the non-diagonality
and non-stationarity of the tensor of inertia of bodies relative to the basic
coordinate system, which determine:
• occurrence of inconsistency of the main central axes of inertia of the
system of bodies with the axes associated with the center of mass of the
AMR platform by the basic CX c Y c Z c coordinate system,
• inconsistency of control actions with the directions of the main central
axes of inertia of the system,
• interdependence of control channels.
A mathematical model of controlled motion of AMR with M and an algo-
rithm for solving the problem of dynamics by the Runge–Kutta–Fehlberg
method are presented. Software has been developed for identifying the cur-
rent dynamic state of AMR with M and assessing the mutual connection of
control channels.
 References 257

An algorithm for adaptive control of AMR with M has been developed,

which provides for the simultaneous study and assessment in real time of the
dynamic characteristics of AMR (identification problems) and determination
of control parameters from the obtained model (synthesis problem). The
proposed control algorithm makes it possible to equip work with means of
automatic learning (self-learning) and adaptation, which turn their control
system into an extremely flexible tool for organizing purposeful behavior
under conditions of uncertainty.
The research carried out allows to get close to the creation of an adaptive
multi-connected control system for AMR with M to increase its survivability
and efficiency during autonomous operation in extreme conditions. Further
research is aimed at developing predictive control [29], and control of the
AMR group [30].

Acknowledgments
The results presented in the article were obtained by the authors in the
course of research on the topic 0120U103294 “Universal anthropomorphic
manipulator” and on the topic 0119U101151 “Applied research in mechanics
and mechatronics.”

References
[1] N. Ashchepkova, S. Ashchepkov, S. Kapera, ‘Dynamics of transport
robot model during the turns’, Science and Education in New Dimen-
sion, pp. 26–29, VI(19), Issue 171, 2018. (in Ukrainian)
[2] B. Siciliano, O. Khatib, ‘Handbook of Robotics’, pp. 799–825, Springer,
2008.
[3] S. Tkachev, ‘Stabilization of nonminimum-phase multi-input affine
system’, Science and Education. Scientific periodical of the Bauman
MSTU, #8, 2012.
[4] O. Andrianova, ‘Path following simulation of wheleed vehicle’, Science
and Education. Scientific periodical of the Bauman MSTU, #10, 2011.
[5] R. Gilimyanov, A. Pesterev, L. Rapoport, ‘Motion control for a wheeled
robot following a curvilinear path’, Journal of Computer and Systems
Science International, pp. 987–994, Vol. 47, #6, 2008.
[6] A. Kanatikov, T. Kasatkina, ‘Features of transition to path coordinates
in a problem of path stabilization’, Science and Education. Scientific
periodical of the Bauman MSTU, #7, 2012.
258 Analysis of the Dynamics and Controllability of an Autonomous Mobile Robot

[7] S. Bai, L. Zhou, G. Wu, ‘Manipulator Dynamics’, in book: Handbook

of Manufacturing Engineering and Technology, pp. 799–825, Springer,
2015 DOI: 10.1007/978-1-4471-4670-4_91
[8] M. Orsag, C. Korpela, P. Y. Oh, S. Bogdan, ‘Aerial Manipulator
Dynamics’, in book: Aerial Manipulation, pp. 799–825, Springer, 2018,
DOI: 10.1007/978-3-319-61022-1_5
[9] E. Jurevich, ‘Basis of robot’, 304 p., BHV - Petersburg, St. Petersburg,
2017. (in Russian)
[10] S. Kolyubin, ‘Dynamics of the robots systems’, 117 p. Publishing house
of ITMO University, St. Petersburg, 2017. (in Russian)
[11] N. Ashchepkova, ‘Mathcad in the kinematic and dynamic analysis of
the manipulator’, Eastern-European Journal of Enterprise Technologies,
pp. 54–63, Vol. 5/7 (77), 2015. DOI: 10.15587/1729-4061.2015.51105
(in Russian)
[12] N. Ashchepkova, ‘Divising a method to analyze the current state of
the manipulator workspace’, Eastern-European Journal of Enterprise
Technologies, pp. 63–74, Vol. 1/7 (109), 2021. DOI: 10.15587/1729-
4061.2021.225121
[13] N. Ashchepkova, ‘Determination of optimal configurations of an
antropomorphic manipulator with six degress of mobility’, Herald of
the National Technical University “KhPI”. Series of: Informatics and
Modeling. Vol. 28(1353), pp. 94–107, December 2019, DOI: 10.20998
/2411-0558.2019.28.01, (in Russian)
[14] S. Lloyd, R. Irani, M. Ahmadi, ‘A numeric derivation for fast regressive
modeling of manipulator dynamics’, Mechanism and Machine The-
ory,#156:104149, February 2021. DOI: 10.1016/j.mechmachtheory
.2020.104149
[15] J. Khurpade, S.S. Dhami, s. S. Banwait, ‘A Virtual Model of 2D Planar
Manipulator Dynamics’ International Conference on Smart Systems and
Inventive Technology (ICSSIT), December 2018, DOI: 10.1109/ICSS
IT.2018.8748674
[16] S.X. Tian, S.Z. Wang, ‘Dynamic Modeling and Simulation of a Manip-
ulator with Joint Inertia’, International Symposium on Information
and Automation, ISIA 2010: Information and Automation, pp. 10–16,
2010.
[17] N. Ashchepkova, ‘Analysis of vibrations round centre-of-mass system
the ‘Small space vehicle with a manipulator’, Herald of the Dnepr
National University. Series of: Space-rocket, pp. 11–17, Vol. 4/1 (13),
2009. ISSN 9125 0912, (in Russian)
 References 259

[18] S. Dubowsky, E. Papadopoulos, ‘The Kinematic, Dynamic and Control

of Free Flying Systems’, IEEE Transactions on Robotic and Automa-
tion, Vol. 9(5), 1993.
[19] P. Efimova, A. Shimanchuk, ‘Design of motion of space manipulation
robot‘, Problems of mechanics and contral: Nonlinear dynamic systems,
pp. 20–30, Vol. 46, 2014. (in Russian)
[20] L. Love, J. Jansen, F. Pin, “On the modeling of robots operating
onships”, in Robotics and Automation, Proc. In IEEE International
Conference (ICRA’04), pp. 2436–2443, 2004.
[21] L. Love, J. Jansen, F. Pin, ‘Compensation of wave-induced motion
and force phenomena for ship-based high performance robotic and
human amplifying systems’, Technical Report ORNL/TM-2003/233.
Oak Ridge National Laboratory, Oak Ridge, Tenn, 2003.
[22] P. J. From, V. Duindam, J. T. Gravdahl, ‘Modeling and motion plan-
ning for mechanisms on a non-inertial base’, Proc. In IEEE Interna-
tional Conference on Robotics and Automation - 2009, pp. 3320–3326,
2009.
[23] E. Sadraei, M. M. Moghaddam, ‘On a Moving Base Robotic Manipula-
tor Dynamics’, International Journal of Robotics, Vol. 4, No. 3, October
2015, pp. 66–74, 2015.
[24] C. M. Wronka, M. W. Dunnigan, ‘Derivation and analysis of a dynamic
model of a robotic manipulator on a moving base’, Robotics and
Autonomous Systems, Vol. 59, pp. 758–769, 2011.
[25] W. Vereecken, et. al., ‘Energy Efficiency in thin client solutions’, ICST
Int. Conf. on Networks for Grid Applic., Athens, 2009.
[26] N. Ashchepkova, A. Zbrutsky, ‘Modeling the dynamics of an
autonomous mobile robot with a manipulator’, Herald of the National
Technical University “KhPI”. Series of: Informatics and Modeling. Vol.
2(4), pp. 34–45, December 2020, DOI: 10.20998/2411-0558.2020.01.0
3. (in Ukrainian)
[27] N. Ashchepkova, O Bulaniy,. ‘A decision of equalizations of dynamics
of space vehicle is with the undiagonal tensor of inertia’, Question of
optimization of calculations, Proceedings of the international science
and technology conference, NAS of Ukraine, pp. 12–15, Kyiv, October
6–8, 1997. (in Russian)
[28] M. Brdni, Adaptive control of manipulator robots: Author’s thesis,
Moscow, 18 p., 1993.
260 Analysis of the Dynamics and Controllability of an Autonomous Mobile Robot

[29] V.M. Kuntsevich, et. al (Eds). Control Systems: Theory and Applica-
tions. Series in Automation, Control and Robotics, River Publishers,
Gistrup, Delft, 2018.
[30] Y.P. Kondratenko, V.M. Kuntsevich, A.A. Chikrii, V.F. Gubarev,
(Eds). Advanced Control Systems: Theory and Applications. Series in
Automation, Control and Robotics, River Publishers, Gistrup, 2021.
 11
Safe Navigation of an Autonomous Robot
in Dynamic and Unknown Environments

Yuriy P. Kondratenko1 , Arash Roshanineshat2 ,

and Dan Simon3
1 PetroMohyla Black Sea National University, Intelligent Information
Systems Dept., Mykolayiv, 54003, Ukraine
2 The University of Arizona, Electrical & Computer Engineering Dept.,

Tucson, AZ, 85721, USA

3 Cleveland State University, Electrical Engineering and Computer Science

Dept., Cleveland, OH, 44115, USA

E-mail: [email protected]; [email protected];
[email protected]

Abstract
For robots to become an efficient means for performing tasks, they need to
navigate safely and effectively without supervision. This chapter is devoted
to the problem of fuzzy control system design for mobile robots operating in
dynamic environments with obstacles. The decision-making method, based
on the transformation of radar-sensor information to fuzzy set “Obstacles”
and robot’s heading course to fuzzy set “Direction to Target Point,” is
considered. We also address the safety aspects of the proposed approach.
The different scenarios of the working environment (with static and
dynamic obstacles), the adjusted speed of the autonomous robot, and pecu-
liarities of “robot-obstacle” interaction are the issues for discussion. The
simulation results confirm the efficiency and safety of the robot’s autonomous
navigation in dynamic and unknown environments.

261
262 Safe Navigation of an Autonomous Robot in Dynamic

Keywords: The autonomous robot, unknown environment, static and

dynamic obstacles, fuzzy set, membership function, adjusting speed, safety

11.1 Introduction
Service robots acting in uncertain environments have become very popular in
the last few years [1–5]. Many robots and robotic systems successfully work
in hospitals, office buildings, department stores, museums, enterprises, etc.
[6–8].
For robots to become capable team members and helpful assistants,
especially in dynamic human-populated environments, they need to navigate
efficiently and safely [9–12] in the target area.
In many cases, the efficiency of moving vehicles or robot missions
depends on the properties of the sensor and control systems in providing the
obstacles avoiding [13–20] (spacecraft, ships, crewless underwater vehicles,
etc.). The target area, especially for mobile and underwater robots, often
consists of static and dynamic obstacles. Thus, it is necessary to consider
various environmental uncertainties. The obstacle avoidance with a priori
unknown parameters is the peculiar feature of the control systems for above
mentioned autonomous vehicles and robots.
Intelligent systems based on fuzzy sets and fuzzy logic [3, 7, 11, 12, 21-
34] can be a good solution for this problem, especially when a system’s
mathematical model is either unavailable or too complex. Many publications
[1, 3, 9, 11, 12, 35–45] are devoted to robot trajectory planning in the
working environment with obstacles. However, this problem still requires
developing new models and algorithms that optimize and improve the process
of automatic path planning in real-time.
The authors of this chapter develop and investigate intelligent algorithms,
providing the efficient planning and optimization of the robot’s trajectory in
uncertain and dynamic environments. The proposed algorithm is based on the
representation of radar-distance observations and direction to the target point
by specially constructed fuzzy sets with adjusted parameters of membership
functions according to the current “robot-obstacle” situations.
This chapter pays special attention to an autonomous robot’s efficient and
safe navigation in a dynamic or a priori unknown environment. Efficient
navigation means the ability of the robot to reach a target point from a
starting position in a priori unknown environment, and the safety measure
corresponds to the number of the robot collisions with dynamic obstacles
during its navigation to the target point.
 11.2 Related Works 263

11.2 Related Works

Robot motion planning in dynamic environments has recently received sub-
stantial attention due to the advent of autonomous cars and growing interest
in social, service, and assistive robots.
Deepu et al. [46] focus on providing a path for a robot to ease movement,
detecting and avoiding obstacles in an environment using a single camera and
a laser source. In [47] authors propose a potential hybrid field, which can be
computed in real-time, to navigate a robot in the dynamic environment with
50 randomly moving obstacles. A fuzzy-inference system with an acceler-
ate/brake module is developed in [48] for real-time navigation of autonomous
underwater vehicles in both static and dynamic three-dimensional environ-
ments with automatically avoiding the dynamic obstacles using sonar model,
virtual acceleration, and velocity in both horizontal and vertical planes. This
chapter [49] presents a fuzzy truck control system for obstacle avoidance
with a reasonably good trajectory, using 33 fuzzy inference rules for an
autonomous vehicle’s steering control and 13 rules for speed control. In
[50], two fuzzy logic controllers for an autonomous vehicle’s steering and
velocity control consist of seven control modules. It provides (a) motion to
the target point, (b) vehicle’s final orientation, (c) avoiding collision with
obstacles, (d) driving the vehicle through mazes, and (e) control the velocity
of the movement toward the target point in the neighborhood of obstacles
or when the vehicle turns sharp corners. The authors consider in [51] a
reactive strategy for navigating a mobile robot in dynamic a priori unknown
environments densely cluttered with moving and deforming obstacles.
Learning is essential for cognitive robots in adapting the human experi-
ence to intelligent robot control in the natural environment [52–57]. Learning
can assist in future robot decisions for both efficiency and robustness. In
[58], authors propose to form a robot’s experience online and to trans-
fer knowledge among appropriate contexts using the learning paradigm of
inductive logic programming to frame hypotheses. Using first-order logic for
such hypotheses representation is helpful for further reasoning and planning
processes.
Combining artificial neural networks and evolutionary algorithms is an
effective method for providing robustness in different navigation conditions.
An example is an autonomous robot “Khepera” [59], with a sensory-motor
model, based on the neural network and genetic algorithm, with the fitness
evaluation in terms of the navigation performance in a maze course. In [60]
authors present the framework for the navigation and target tracking system
264 Safe Navigation of an Autonomous Robot in Dynamic

for the mobile robot, which uses a Microsoft Xbox Kinect sensor. The main
task of the fuzzy controller is providing control of the robot in obstacle avoid-
ance and target following. The designed multi-robot system [60] can work
autonomously in an outdoor environment. In [61], the various robot motion
planning schemes, including a genetic-fuzzy system, genetic-neural system,
and a conventional potential field approach, have been compared in terms
of traveling time taken by the robot, robustness, adaptability, goal-reaching
capability, and repeatability. Neural networks, genetic algorithms and genetic
programming are augmented with fuzzy logic-based schemes to enhance
artificial intelligence of mobile robots [14, 16, 29, 30, 34, 35, 39, 42, 62–67].
In [68], the authors propose a neuro-fuzzy system architecture for behavior-
based control of a mobile robot in unknown environments. The methodology
of the behavior-based control approach includes two steps: (a) to analyze and
to decompose a complex task based on stimulus-response behavior, (b) to
formulate (quantitatively) each type of behavior with a simple feature by
fuzzy sets and fuzzy rules, and (c) to coordinate conflicts and competition
among multiple types of behavior by fuzzy reasoning.
An integrated representation of the environment with an approximation
of obstacles’ shapes by discs or polygons [69] is the base for collision-free
navigation of a mobile robot in complex dynamic environments with moving
obstacles. The navigation algorithm presented in [69] provides a short robot
path planning through the crowd of moving or steady obstacles.
An efficient stereovision-based motion compensation method for moving
robots, discussed in [70], uses the disparity map and three modules: seg-
mentation, feature extraction, and estimation. In the segmentation module,
the authors propose using extended type-2 fuzzy sets to extract the objects.
Fuzzy logic is used for the robot in [71] to implement the behavior design
and coordination during the realization of the “memory grid” and “minimum
risk method.” The robot can choose the safest region to avoid colliding with
obstacles in different scenarios, particularly in the long wall, large concave,
recursive UU-shaped, unstructured, cluttered, maze-like, and dynamic indoor
environments. This chapter [72] presents the fuzzy-logic controller based on
the Mamdani-type fuzzy inference engine for robot navigation and obstacles
avoidance in a cluttered environment. The fuzzy controller with three inputs
and single output provides safe navigation of the robot in a static environment
considering the accuracy of absolute measurements of its position, obstacle
distances, goal distance, velocity, orientation, and rate of change in its head-
ing angle. This chapter [73] describes a fast and reliable obstacle avoidance
method for ground mobile robots in both outdoor and indoor navigation. The
 11.2 Related Works 265

process realizes two contradictory approaches (non-complex implementation

and human-like smooth steering) and applies in different mobile robotic
systems regardless of used sensors. In [74], a conceptual fuzzy logic approach
is the basis for solving a multi-link robot’s local navigation and obstacle
avoidance problem. The proposed method is the core of developing an
online local navigation system for generating instantaneous collision-free
trajectories.
Thus, the analysis of last publications on autonomous and mobile robots
and their motion control in the environment with obstacles shows [49, 60, 61]
that researchers continue to develop new avoiding obstacles’ algorithms
and new devices’ solutions using well-known and new design methods,
approaches, and methodologies. It deals, first of all, with the different
and specific missions of the mobile robots; the high level of information
uncertainty concerning the nature of obstacles and the character of their
appearance; changes of obstacles’ movement velocities; a large set of the
various geometrical shapes of obstacles; unknown level of the danger in the
case of an accident, etc.
Additional design requirements also stimulate researchers to develop new
approaches, methods, and algorithms for avoiding obstacles by a mobile
robot. Among such most important requirements are:
(a) restricting the design space in the robot’s construction for corresponding
navigation and control devices’ placement (that is a vital requirement for
design processes in the underwater and space robotics);
(b) simplifying the information and signal processing algorithms for embed-
ded robot’s navigation and control systems with providing the trajectory
planning in real-time [75];
(c) reconfiguring the information processing algorithms and corresponding
systems [76] based on the modern field-programmable gate array -
FPGA [63, 77–79];
(d) providing structural-parametrical optimization and high reliability of
the computerized control systems [30, 32, 34, 52, 80–85] for robot
navigation, etc.
The stringent requirements to the robot’s navigation and control systems’
parameters allow realizing efficient trajectory planning in real-time. In
particular, it requires the minimization of such parameters as:
• the response time between the moment of obstacle emergence in the
“vision” zone of the mobile robot and the moment of its recognition by
the robot’s navigation system;
266 Safe Navigation of an Autonomous Robot in Dynamic

• the time of the fuzzy information processing for the calculation and cor-
rection of the robot motion’s course providing the obstacle avoidance;
• the time for calculating and correcting the robot motion’s velocity to
secure the save motion near the obstacle.
Therefore, the actual research task is to modify the existing and create new,
more efficient navigation and control systems algorithms for their successful
implementation in mobile robots and autonomous vehicles.
As discussed above, the artificial intelligence methods, including the
application of fuzzy sets and fuzzy logic, are powerful tools for solving such
problems as planning efficient trajectories for the safe movement of mobile
robots in an uncertain environment with obstacles. Choosing a corresponding
artificial intelligence method for implementation in a robot navigation-control
system (providing efficient fuzzy information processing) requires taking into
account: (a) all specific robot’s environment conditions and peculiarities;
(b) an extensive spectrum of the possible missions for a mobile robot; (c)
the universality for different real working scenarios with static and dynamic
obstacles and other uncertainties.
Thus, it is reasonable to develop new design methods, algorithms, and
models that provide efficient fuzzy information processing. It will simul-
taneously improve the design processes for robots’ navigation-control sys-
tems and enhance the control indicators of their functioning in uncertain
environments with obstacles.
The main idea of this chapter is an investigation of the fuzzy approach for
mobile robots’ trajectory planning in an environment with uncertain obsta-
cles, using proposed by authors fuzzy reasoning algorithm. This algorithm
will provide: (a) the increasing levels of universality for different uncertain
scenarios with obstacles, and (b) simplicity by decreasing the complexity
of the fuzzy information processing and by reducing intricacy of the cor-
responding device’s synthesis process for hardware realization, as well as, (c)
increasing the reliability and (d) decreasing the time for decision making in
different conflict situations.
Finally, let us formulate the aims of this chapter as:
• introducing the fuzzy reasoning algorithm [21–25] instead of fuzzy-rule-
base inference engine for increasing efficiency of robot’s navigation in
the uncertain environment with unknown obstacles;
• the investigation of proposed by authors “radar-fuzzy-set” fuzzy
processing approach, concerning the continuous adjusting of robot
movement parameters and parameters of fuzzy reasoning algorithm,
based on the current information of the radar-sensory system;
 11.3 Problem Statement 267

• developing the structured multi-agent software for modeling and simula-

tion of the different scenarios of robot’s navigation with confirmation of
the efficiency of the developed fuzzy processing algorithms for robot’s
navigation in uncertainty.
The rest of the chapter covers multiple aspects related to the topic under
discussion. Section 11.3 considers a general representation of the proposed
method of fuzzy information processing for robot navigation in the environ-
ment with unknown obstacles, based on the intersection of two dynamic fuzzy
sets, “Environment with obstacles” and “Fuzzy direction to the target point.”
In Section 11.4, the authors provide a more detailed description of the fuzzy-
based path planning for investigation of the main properties of the proposed
robot’s navigation system. Section 11.5 deals with the developed algorithm
for robot speed adjusting in a dynamic environment with obstacles. Section
11.6 represents the modeling and simulation results for the different mobile
robot’s navigation scenarios in an uncertain environment with the radar-
sensory system for obstacles detection. This chapter ends with a conclusion
in Section 11.7.

11.3 Problem Statement

Figure 11.1 shows a sample of a working environment in a two-dimensional
space. The area is a priori unknown to the autonomous robot. SP is the initial
(start) position of the robot, and TP is the target point.
The robot has a 360◦ radar sensor with a limited range Lmax . The problem
is that the robot should autonomously synthesize the non-linear trajectory
for efficient movement from SP to TP, avoiding all the static and dynamic
obstacles and providing (a) the minimum length of the trajectory, (b) the
minimum time for robot motion on the trajectory, and (c) high reliability for
movement in the dynamic working environment.
Online trajectory planning can be done through continuous fuzzy infor-
mation processing of the radar sensor data and identifying the robot’s
direction at the beginning of each motion step on the planning trajectory.
The following section considers the proposed algorithm for such real-time
trajectory planning in detail. It is supposed that the mobile robot’s current
position (coordinates) SP and the coordinates of the target point TP in the
target area are known or can be measured at every step of the trajectory
planning.
268 Safe Navigation of an Autonomous Robot in Dynamic

Figure 11.1 The scenario of mobile robot’s working environment with obstacles

Table 11.1 Table of Notations

ϕi The angle of the radar beam
Δϕ Step of the radar beam
Li Distance to the object in a ϕi direction
Lmax Maximum distance of radar beam
E
∼
Radar output fuzzy set
μ∼
E(ϕi ) Radar output membership function
α The angle between robot position and target point
G
∼
Our introduced rectangular fuzzy set
G (ϕi )
μ∼ Chosen rectangular membership function
R
∼
Resulted fuzzy set in intersection stage
μR
∼
(ϕi ) Resulted membership function from intersecting rectangular
membership function and radar output
ψ Correct angle to choose by the robot
 11.4 Fuzzy based Path Planning 269

11.4 Fuzzy based Path Planning

The robot can detect the obstacles within its radar sensor [3, 12, 86], scanning
the working space by a beam of a limited length Lmax . Such beam is turning
in a positive direction ϕi with a discrete step Δϕ = 2πn :

2π
ϕi+1 = ϕi + Δϕ = (i + 1) , i = 0, 1, 2, ...n − 1,
n
where n is the number of discrete steps for a scanning beam; Li is a distance
between the robot and the nearest obstacle at every ith step ϕi of the scanning
beam, Li /Lmax , Li ∈ [0, 1].
Figure 11.2 illustrates the circular scan (radar-beam-diagram) for the
configuration of the obstacles Obs2 and Obs3 at the mobile robot’s trajectory
point A (Figure 11.1). The distance Li , (i = 1, 2, ..., n) between the robot

Figure 11.2 The radar-beam-diagram of the robot sensor system in point A on the motion
trajectory, presented in Figure 11.1
270 Safe Navigation of an Autonomous Robot in Dynamic

(point 0 in the center of the circle diagram, Figure 11.2) and any obstacle on
the corresponding beam is presented in the relative form Li /Lmax ∈ [0, 1].
The proposed algorithm for fuzzy information representation and process-
ing consists of the following stages:
Stage 1. Scanning radar area. The interpretation of the scanned results of
the robot’s visible zone (domain) {L0 , L1 , ..., Li , , ..., Ln } for the working
environment with detected obstacles as the fuzzy set E. ∼
Such fuzzy set E∼
(for fuzzy interpretation of the radar information shown
in Figure 11.2) is shown in Figure 11.3, considering the following dependence
Li
μE
∼
(ϕi ) = , (11.1)
Lmax
where μE∼
(ϕi ) is a value of the membership function μE ∼
E,
of the fuzzy set ∼
which corresponds to the value of beam’s angle position ϕi , included to the
support S (E)
∼
[87] of the fuzzy set E:
∼

∀ϕi : ϕi ∈ S (∼ E) = supp (∼ E) = ϕi : μE ∼
(ϕ i ) > 0, ϕi ∈ [0, 2π] .

If any obstacle is absent on the ith scanning beam, the value of member-
Li
ship function μ∼E (ϕi ) = Lmax = 1, because Li = Lmax . If the scanning beam
E (ϕi ) = Lmax ∈ [0, 1] , because Li < Lmax . If
Li
meets the obstacle, then μ∼
the robot contacts a corresponding obstacle or robot situated at the minimal
admissible distance Lmin to the obstacle (and the further movement in the
current direction is impossible), then μE ∼
(ϕi ) = LLmaxi
= 0, as Li = 0 or
Li ≤ Lmin .
It is evident that the membership function of the fuzzy set E ∼
has a
rectangular form for the case if no obstacle is in the radar’s visible zone

∀ϕi ∈ [0, 2π] : μE

∼
(ϕi ) = 1.

It is possible to calculate the recommended value of the robot course

(motion direction) ϕ∗ , that has the optimal distance from the visible obstacles
using any defuzzification algorithm [88–90] for the interpreted fuzzy set
(target area). The value of the recommended course ϕ∗ = 0.99π , presented
in Figure 11.3, was calculated using the gravity center method for the defuzzi-
fication of the fuzzy set E
∼
(Figure 11.3). This recommended value ϕ∗ secures
only the safe course, but it is not related to the final target position of the robot
after its accomplishing mission.
 11.4 Fuzzy based Path Planning 271

Figure 11.3 The fuzzy set E,

∼
which illustrates the proposed fuzzy interpretation of the radar-
beam-diagram, presented in Figure 11.2

Stage 2. Introducing a suitable membership function. This stage deals with

the interpretation of the desired (straight directed) course α to the target
point TP from the current position of the mobile robot by corresponding
fuzzy set G∼ “Fuzzy direction to the target point” with membership function
μG∼
(ϕi ) , ϕi ∈ [0, 2π], where μG ∼
(ϕi ) = 1 in the condition α − c ≤ ϕi ≤
α + c.
Different types and shapes of membership functions can present the fuzzy
set G
∼
and its membership function μ∼ G (ϕi ). We consider the rectangular
membership function to represent the fuzzy set G ∼ in the following form:

1 if α − c ≤ ϕi ≤ α + c
μG (ϕ i ) = (11.2)
∼ 0 otherwise,
where c is a parameter of membership function μ∼ G (ϕi ), which influences the
wideness of the rectangular membership function.
The membership function μG ∼
(ϕi ) presents (Figure 11.4) the example of
the fuzzy set G
∼
for the desired course α = 0 and |c| < 1.
Stage 3. Interacting membership function and radar data. We can find
at this stage the resulting fuzzy set R, ∼
which corresponds to the fuzzy
representation of the mobile robot’s path with securing direction to the target
point and avoiding obstacles. It is possible to calculate the parameters of this
fuzzy set R
∼
as intersection of two normal fuzzy sets G ∼
and E:
∼

R
∼
=G∼
E,
∼
(11.3)
272 Safe Navigation of an Autonomous Robot in Dynamic

Figure 11.4 The fuzzy set G

∼
with rectangular membership function (11.2)


where is a fuzzy intersection operator, for example, t-norm operator [87,
90, 91, 92].
The membership function μR∼
(ϕi ) of the fuzzy set R
∼
can be calculated
using the algorithm

R (ϕi ) = μ∼
μ∼ G (ϕi ) μ∼
E (ϕi ) . (11.4)
Usually, a fuzzy set with a membership function μR ∼
(ϕi ) has a one-
interval or multi-intervals core [87] if an obstacle exists in the radar range
of the robot.
Figure 11.5 illustrates the resulting fuzzy set R ∼
with membership function
μR∼
(ϕi ) for interpreting the robot’s heading course ψ as an intersection of the
two fuzzy sets G ∼
and E ∼
according to (11.4).
The shape of the membership function μ∼ R (ϕi ) mostly depends on the
shapes of the membershipfunctions μG ∼
(ϕi ) and μE∼
(ϕi ), but the choice of
the intersection operator ( ) also influences the final form.
It is possible to use different types of t-norm operators ( ) as intersection
operators from two families [87] of non-parameterized and parameterized
t-norm operators. All t-norm operators satisfy the special conditions of
t-norm: mapping space, zeroing, the identity of unity, commutativity, union,
monotonicity.
For example, we can use the non-parameterized minimum intersection
operator (MIN-operator) for the calculation of the values μ∼ R (ϕi ) according
to (11.4) in the following way:
  
μR∼
(ϕ i ) = μ G
∼
(ϕ i ) μ E
∼
(ϕ i ) = M IN μ G
∼
(ϕ i ) , μE
∼
(ϕ i ) . (11.5)
 11.4 Fuzzy based Path Planning 273

Figure 11.5 The resulting fuzzy set R∼

with one-interval core for the determination of mobile
robot’s heading course according to (11.3), (11.4)

Figure 11.6 The multimodal resulting fuzzy set R

∼
with multi-intervals core

At the same time, it is possible to use (alternatively) in (11.4) the different

parameterized mean operators as intersection operators [87] (which are not
satisfying with special t-norm conditions).
Figure. 11.6 shows the membership function μR ∼
(ϕi ) with the multi-
intervals core of the resulting fuzzy set R.
∼
The next step is finding sections where there are no obstacles. Values
μR∼
(ϕi ) = 1 mean that there is no obstacle in the direction ϕi . In Figure 11.6,
highlighted segments with (1) and (2) show no obstacles in these areas. The
robot should choose a segment that has less probability of collision with an
obstacle.
274 Safe Navigation of an Autonomous Robot in Dynamic

First, the robot finds all sections where the radar distance value is more
than a threshold for finding a segment with less probability. Since there are
measurement noises, radar may not have the exact value of 1.

a) No Obstacles, if
μ∼ (ϕi ) ≤ threshold
Determining Obstacles in μR
∼
(ϕi ) = R
b) Obstacles, if
R (ϕi ) > threshold
μ∼

Then the robot chooses the segment which is the longest. Figure 11.6 shows
that segment (1) is longer than segment (2), so the robot will select a part (1).
Defuzzification of this segment will return the most appropriate direction
for the robot. In some cases, the choice of the defuzzification method is
limited by the necessity to take into account the multi-intervals core of the
resulting fuzzy set ∼R and its membership function μR ∼
(ϕi ) that practically
does not meet in the majority of the Mamdani-type fuzzy inference systems.
The multimodality presence explains that if the direct way to the target point
is blocked for the mobile robot by an obstacle, there can be some alternate
route concerning the shape of the specified obstacle.
Stage 4. Defuzzification. The last stage is the determination of the crisp value
of the robot’s direction ψ by defuzzification of the resulting fuzzy set R
∼
with
membership function μR ∼
(ϕi ), presented as an example in Figure 11.5:
'  (
ψ = Arg Def uzzif ication μR ∼
(ϕi ) (11.6)

Different defuzzification algorithms can be used [87–90], particularly the

first maxima, the last maxima, the middle maxima, the center of gravity, the
center of sum, the bi-sector, the height method.
The current value of the robot’s heading course ψ should be calculated
in real-time by the robot during its navigation to the target point TP avoiding
any obstacles taking into account the changeable character of both fuzzy sets
G
∼ and E.∼
Thus as mathematical modeling shows, situations are frequent when two
or more course angles have an equivalent grade of the membership to the
R. The inadequacy of such commonly used defuzzification methods
fuzzy set ∼
as a center of gravity or bisector for a case of multimodal membership
functions is shown in [75].
Figure 11.7 shows the highlighted area for defuzzification of the resulting
fuzzy set ∼R with the multi-intervals core.
 11.6 Experiments and Evaluations 275

Figure 11.7 The highlighted area for defuzzification

11.5 Adjusting Speed

Some situations (in dynamic environments with obstacles and real-time con-
trol of mobile robots movement) need increasing or decreasing robot motion
speed (speed control).
In general, the mobile robot has enough time for re-planning its path if
the robot detects a static obstacle. Thus, we will consider the situations with
changing the speed of the robot only if dynamic obstacles [3, 12, 86, 93, 94]
are detected.
Figure 11.8 shows different divisions we have considered for the speed of
the autonomous robot.
There are five-speed parts in total (Figure 11.8) according to the different
descriptions in Table 11.2.

11.6 Experiments and Evaluations

We have experimented with our method on four maps and six times for each
of them. In three of the experiments, safe navigation was activated, and in the
other three experiments, safe navigation was disabled. Numbers (Table 11.3)
are the absolute values of the average of three experiments for both safe and
unsafe experiments.
Dynamic obstacles have pink and static obstacles have black borders
(Figure 11.9).
The size of the maps is 500x400 pixels, and the radar radius is 30 pixels.
Since there is no dynamic obstacle in Map 1 (Figure 11.9), the robot could
reach the target point without colliding with any obstacle in safe and unsafe
modes.
276 Safe Navigation of an Autonomous Robot in Dynamic

Figure 11.8 The robot radar division for different speeds

Table 11.2 Table of the robot speed

No. Speed Condition 1 Condition 2
1 Maximum Speed 360 − θ1 > min (ϕi ) > θ1 Lmin(ϕi ) < 0.5
2 High Speed 360 − θ1 > min (ϕi ) > θ1 0.5 < Lmin(ϕi ) < 1
3 Low Speed θ2 < min (ϕi ) < θ2 0.5 < Lmin(ϕi ) < 1
4 Stop θ2 < min (ϕi ) < θ2 Lmin(ϕi ) < 0.5
5 Normal Otherwise

Table 11.3 Result of experiments

The Average Number of The Average Number of
Collisions in Safe Collisions in Unsafe
Navigation Navigation
Map 1 0 0
Map 2 0 8
Map 3 0 16
Map 4 0 33

In Map 2, Map 3, and Map 4 (Figure 11.9), in all three corresponding

safe mode experiments, the robot reached the target point without colliding
with any obstacle, but in unsafe experiments, the robot was hit with dynamic
obstacles (Table 11.3).
 11.7 Conclusion 277

Figure 11.9 Maps of unknown environments used for experiments

11.7 Conclusion
We propose a fuzzy-based approach enabling robots to navigate efficiently
and safely in uncertain environments.
Experimental results show the applicability of the proposed approach
for autonomous robot navigation in dynamic environments where mov-
ing obstacles are avoided efficiently, and the robot manages to reach its
destination.
In our experiments, the autonomous robot chooses a safe path for naviga-
tion. In situations where it needs to avoid collisions with dynamic obstacles,
the robot can adjust its speed to escape from dangerous situations as safely
as possible. We showed that changing rate resulted in having no collisions
between the robot and obstacles.
278 Safe Navigation of an Autonomous Robot in Dynamic

For future work, we would like to extend this work to a multi-robot setting
and investigate the robustness of sensor-based multi-robot control in an
uncertain environment [95–99] using hybrid artificial intelligence methods.

Acknowledgments
The authors thank the Fulbright Program (USA) for supporting Prof. Y. P.
Kondratenko with a Fulbright scholarship and making it possible for this team
to conduct research together in the USA at the Cleveland State University,
Electrical Engineering, and Computer Science Department.

References
[1] M. Derkach, D. Matiuk, I. Skarga-Bandurova, ‘Obstacle Avoidance
Algorithm for Small Autonomous Mobile Robot Equipped with Ultra-
sonic Sensors,’ in IEEE 11th Int. Conf. on Dependable Systems,
Services, and Technologies, DESSERT 2020, pp. 236–241, 2020.
[2] A.N. Tkachenko, N.M. Brovinskaya, Y.P. Kondratenko, ‘Evolutionary
adaptation of control processes in robots operating in non-stationary
environments,’ Mechanism and Machine Theory, Vol. 18, No. 4, pp.
275–278, 1983. DOI: 10.1016/0094-114X(83)90118-0
[3] Y. Kondratenko, G. Khademi, V. Azimi, D. Ebeigbe, M. Abdelhady,
S. Fakoorian, T. Barto, A. Roshanineshat, I. Atamanyuk, D. Simon,
‘Robotics and Prosthetics at Cleveland State University: Modern Infor-
mation, Communication, and Modeling Technologies,’ in Information
and Communication Technologies in Education, Research, and Indus-
trial Applications, ICTERI 2016. Communications in Computer and
Information Science, Vol. 783. Springer, Cham, pp. 133–155, 2017.
DOI: https://doi.org/10.1007/978-3-319-69965-3_8
[4] Y. Kondratenko, L. Klymenko, V. Kondratenko, G. Kondratenko,
E. Shvets, ‘Slip Displacement Sensors for Intelligent Robots: Solu-
tions and Models’, in Proc. of the 2013 IEEE 7th Int. Conf.
on Intelligent Data Acquisition and Advanced Computing Systems
(IDAACS), Berlin, vol. 2, Sept. 12–14, pp. 861–866, 2013. DOI:
10.1109/IDAACS.2013.6663050
[5] U. Patel et al., ‘Beam: A Collaborative Autonomous Mobile Service
Robot,’ in Proceedings of the AAAI Fall Symposium on Artificial
Intelligence for Human-Robot Interaction (AI-HRI), 2017.
 References 279

[6] Z. Li and Z. Huang, ‘Design of a type of cleaning robot with ultrasonic,’

in Journal of Theoretical and Applied Information Technology, Vol. 47,
No. 3, pp. 1218–1222, 2013.
[7] Y.P. Kondratenko, ‘Robotics, Automation and Information Systems:
Future Perspectives and Correlation with Culture, Sport and Life Sci-
ence,’ in Decision Making and Knowledge Decision Support Systems,
Lecture Notes in Economics and Mathematical Systems, Vol. 675, A.
M. Gil-Lafuente, C. Zopounidis, Eds. Springer International Publishing
Switzerland, pp. 43–56. 2015. DOI: 10.1007/978-3-319-03907-7_6
[8] M.O. Taranov, et al., ‘Models of Robot’s Wheel-Mover
Behavior on Ferromagnetic Surfaces,’ in International Journal
of Computing, Vol.17, Issue 1, Open Access, pp. 8–14, 2018.
DOI: https://doi.org/10.47839/ijc.17.1.944
[9] M. D’Arcy, P. Fazli, D. Simon, ‘Safe Navigation in Dynamic, Unknown,
Continuous, and Cluttered Environments,’ in Proceedings of the IEEE
International Symposium on Safety, Security, and Rescue Robotics
(SSRR), 2017.
[10] M. Taranov, et al., ‘Simulation of Robot’s Wheel-Mover on
Ferromagnetic Surfaces,’ in IEEE 9th Int. Conf. IDAACS,
Bucharest, Romania, vol. 1, Sept. 21–23, pp. 283–288, 2017. DOI:
10.1109/IDAACS.2017.8095091
[11] D. Driankov, A. Saffiotti (Eds.), ‘Fuzzy Logic Techniques for
Autonomous Vehicle Navigation,’ Physica, 2013.
[12] Y.P. Kondratenko, S.A. Sydorenko, ‘Fuzzy-based trajectory planning in
environment with obstacles’, in Proceedings of East-West Fuzzy Col-
loquium 2006, 13th Zittau Fuzzy Collquium, Institut fur Prozesstech-
nik, Prozessautomatisierung und Messtechnik, Zittau, Sept. 13–15, pp.
40–48, 2006.
[13] Y.P. Kondratenko, V.L. Timchenko, ‘Increase in Navigation Safety by
Developing Distributed Man-Machine Control Systems,’ in Third Inter-
national Offshore and Polar Engineering Conference, Singapore, Vol. 2,
pp. 512–519, 1993, Paper Number: ISOPE-I-93-173.
[14] Y.P. Kondratenko, J. Rudolph, O.V. Kozlov, Y.M. Zaporozhets, O.S.
Gerasin, ‘Neuro-fuzzy observers of clamping force for magnetically
operated movers of mobile robots,’ in Technical Electrodynamics, No.
5, pp. 53–61, 2017 (in Ukrainian). DOI: 10.15407/techned2017.05.053
[15] V.L. Timchenko, et al. ‘Robust stabilization of Marine Mobile Objects
on the Basis of Systems with Variable Structure of Feedbacks,’ in
Journal of Automation and Information Sciences, Vol. 43, No. 6. –
280 Safe Navigation of an Autonomous Robot in Dynamic

New York: Begel House Inc., pp. 16–29, 2011. DOI: 10.1615/JAutomat-
InfScien.v43.i6.20
[16] R. Duro, Y. Kondratenko (Eds.), ‘Advances in Intelligent Robotics and
Collaborative Automation’, in Book Series: River Publishers Series in
Automation Control and Robotics, 2015. DOI: https://doi.org/10.13052
/rp-9788793237049
[17] O.S. Gerasin, et al. ‘Remote IoT-based control system of the mobile
caterpillar robot,’ in 16th International Conference on ICT in Educa-
tion, Research and Industrial Applications. Integration, Harmonization,
and Knowledge Transfer, Volume I: Main Conference, ICTERI 2020,
Kharkiv, CEUR Workshop Proceedings, Vol. 2740, pp. 129–136, 2020.
CEUR-WS.org/Vol-2740/20200129.pdf
[18] T. Huntsberger, A. Hrand, E. Baumgartner, S. Paul, ‘Behavior-based
control systems for planetary autonomous robot outposts,’ in Aerospace
Conference Proceedings, 2000 IEEE, vol. 7, IEEE, pp. 679–686, 2000.
[19] D. Simon, ‘Neural Networks for Optimal Robot Trajectory Planning,’ in
E. Fiesler, R. Beale (Eds), Handbook of Neural Computation, Institute
of Physics Publishing, pp. G2.5:1–8, 1997.
[20] T. Huntsberger, ‘Biologically inspired autonomous rover control,’ in
Autonomous Robots 11, No. 3, pp. 341–346, 2001.
[21] L.A. Zadeh, ‘Fuzzy sets. Information and control’, No. 8, pp. 338–353,
1965.
[22] L.A. Zadeh, ‘Outline of a new approach to the analysis of complex
systems and decision processes,’ in IEEE Transactions on Systems, Man
and Cybernetics, Vol. 3, No. 1, pp. 28–44, 1973.
[23] L.A. Zadeh, ‘Inference in fuzzy logic,’ in IEEE Proceedings, Vol. 68,
pp. 124–131, 1980.
[24] L.A. Zadeh, ‘From computing with number to computing with words –
from manipulations of measurements to manipulation with perceptions,’
in IEEE Transactions on Circuits and Systems Part I, Vol. 4, no. 3,
pp. 105–109, 1999.
[25] J. Kacprzyk, ‘Multistage Fuzzy Control: A Model-Based Approach to
Control and Decision-Making’, in Wiley, Chichester, 1997.
[26] Y.P. Kondratenko, N.Y. Kondratenko, ‘Reduced library of the soft com-
puting analytic models for arithmetic operations with asymmetrical
fuzzy numbers,’ in Alan Casey (Ed), Soft Computing: Developments,
Methods, and Applications. Series: Computer Science, technology
and applications, NOVA Science Publishers, Hauppauge, New York,
pp. 1–38, 2016.
 References 281

[27] W. Lodwick, J. Kacprzhyk (Eds), ‘Fuzzy Optimization,’ in STUDFUZ

254, Berlin, Heidelberg: Springer-Verlag, 2010.
[28] R. R. Yager, D. P. Filev, ‘Unified structure and parameter identification
of fuzzy models,’ in Systems, Man and Cybernetics, Vol. 23, issue 4,
1993.
[29] V.M. Kuntsevich et al. (Eds), ‘Control Systems: Theory and Appli-
cations,’ in Book Series in Automation, Control and Robotics, River
Publishers, Gistrup, Delft, 2018. ISBN: 9788770220248
[30] Y.P. Kondratenko, A.V. Kozlov, ‘Parametric optimization of fuzzy con-
trol systems based on hybrid particle swarm algorithms with elite
strategy,’ in Journal of Automation and Information Sciences 51(12),
pp. 25–45, 2019. DOI: 10.1615/JAutomatInfScien.v51.i12.40
[31] V. Shebanin et al., ‘Application of Fuzzy Predicates and Quanti-
fiers by Matrix Presentation in Informational Resources Modeling,’ in
Perspective Technologies and Methods in MEMS Design: Proceed-
ings of the International Conference MEMSTECH-2016. Lviv-Poljana,
Ukraine, April 20–24, pp. 146–149, 2016. DOI: 10.1109/MEM-
STECH.2016.7507536
[32] Y. Kondratenko, D. Simon, ‘Structural and parametric optimization of
fuzzy control and decision making systems,’ in Recent Developments
and the New Direction in Soft-Computing Foundations and Appli-
cations. Studies in Fuzziness and Soft Computing, Zadeh L., Yager
R., Shahbazova S., Reformat M., Kreinovich V. (eds), Vol. 361, pp.
273–289, 2018. Springer, Cham. DOI: https://doi.org/10.1007/978-3-
319-75408-6_22
[33] A. Gozhyj et al., ‘The method of web-resources management under
conditions of uncertainty based on fuzzy logic,’ in Proceedings of the
2018 IEEE 13th International Scientific and Technical Conference on
Computer Sciences and Information Technologies, CSIT 2018, 11–14
Sept. 2018, Lviv, Ukraine, pp. 347–352, 2018. DOI: 10.1109/STC-
CSIT.2018.8526761
[34] Y.P. Kondratenko, A.V. Kozlov ‘Generation of Rule Bases of Fuzzy Sys-
tems Based on Modified Ant Colony Algorithms’, in Journal of Automa-
tion and Information Sciences, Vol. 51, Issue 3, New York: Begel House
Inc., pp. 4–25, 2019. DOI: 10.1615/JAutomatInfScien.v51.i3.20
[35] K.C. Ng, M.M. Trivedi, ‘A neuro-fuzzy controller for mobile robot nav-
igation and multirobot convoying’, in Systems, Man, and Cybernetics,
Part B: Cybernetics, IEEE Transactions on 28, no. 6, pp. 829–840, 1998.
282 Safe Navigation of an Autonomous Robot in Dynamic

[36] A. Ollero, A. García-Cerezo, J. Martínez, A. Mandow, ‘Fuzzy tracking

methods for mobile robots,’ in Applications of fuzzy logic: Towards
high machine intelligence quotient systems 9, pp. 347–364, 1997.
[37] P. Pirjanian, M. Matarić. ‘Multiple objectives vs. fuzzy behavior coor-
dination,’ in Fuzzy logic techniques for autonomous vehicle navigation,
Physica-Verlag HD, pp. 235–253, 2001.
[38] A. Saffiotti et al., ‘Using fuzzy logic for mobile robot control,’ in
Practical applications of fuzzy technologies. Springer US, pp. 185–205,
1999.
[39] Y. Kondratenko, E. Gordienko, ‘Neural Networks for Adaptive Control
System of Caterpillar Turn,’ in Annals of DAAAM for 2011 & Pro-
ceeding of the 22nd Int. DAAAM Symp. Intelligent Manufacturing and
Automation, (20–23 Oct. 2011, Vienna, Austria), Published by DAAAM
International, Vienna, Austria, pp. 305–306, 2011.
[40] Y. Kondratenko, E. Shvets, O. Shyshkin, ‘Modern Sensor Systems of
Intelligent Robots Based on the Slip Displacement Signal Detection,’
in Ann. of DAAAM for 2007 & Proc. of the 18th Int. DAAAM
Symp. Intelligent Manufacturing and Automation. Vienna, pp. 381–382,
2007.
[41] A. Das, R. Fierro, V. Kumar, B. Southall, J. Spletzer, C. Taylor,
‘Real-time vision based control of a nonholonomic mobile robot,’ in
IEEE Intl. Conf. Robotics and Automation (ICRA 2001), Seoul, Korea,
pp. 1714–1719, 2001.
[42] J. Mbede, X. Huang, M. Wang, ‘Robust neuro-fuzzy sensor-based
motion control among dynamic obstacles for robot manipulators,’ in
Fuzzy Systems, IEEE Transactions on 11, no. 2, pp. 249–261, 2003.
[43] A. Saffiotti, ‘Fuzzy logic in autonomous navigation,’ in Fuzzy Logic
Techniques for Autonomous Vehicle Navigation. Physica-Verlag HD,
pp. 3–24, 2001.
[44] H. Surmann, L. Peters, ‘Robot with Fuzzy Controlled Behaviour,’ in
Fuzzy Logic Techniques for Autonomous Vehicle Navigation, Physica-
Verlag HD, pp. 343–365, 2001.
[45] G. Fu, A. Menciassi, P. Dario, ‘Development of a low-cost active 3D
triangulation laser scanner for indoor navigation of miniature mobile
robots,’ in Robotics and Autonomous Systems, Vol. 60, Issue 10,
pp. 1317–1326, 2012.
[46] R. Deepu, B. Honnaraju, S. Murali, ‘Path generation for robot navi-
gation using a single camera,’ in Procedia Computer Science, Vol. 46,
pp. 1425–1432, 2015.
 References 283

[47] S. Ratering, M. Gini, ‘Robot navigation in a known environment with

unknown moving obstacles,’ in Autonomous Robots, 1, pp. 149–165,
1995.
[48] B. Sun, D. Zhu, L. Jiang, S.X. Yang, ‘A novel fuzzy control algorithm for
three-dimensional AUV path planning based on sonar model,’ in Journal
of Intelligent & Fuzzy Systems 26, pp. 2913–2926, 2014.
[49] D.H. Kim, K.B. Kim, E.Y. Cha, ‘Fuzzy truck control scheme for obsta-
cle avoidance,’ in Neural Computing & Applications, 18, pp. 801–811,
2009.
[50] N.E. Hodge, L.Z. Shi, M.B. Trabia, ‘A distributed fuzzy logic controller
for an autonomous vehicle,’ in Journal of Robotic Systems 21(10),
pp. 499–516, 2004.
[51] A.S. Matveev, M.C. Hoy, A.V. Savkin, ‘A globally converging algorithm
for reactive robot navigation among moving and deforming obstacles,’
in Automatica, Vol. 54, pp. 292–304, April 2015.
[52] Y.P. Kondratenko, N.Y. Kondratenko, ‘Soft Computing Analytic Models
for Increasing Efficiency of Fuzzy Information Processing in Decision
Support Systems,’ in Decision Making: Processes, Behavioral Influ-
ences, and Role in Business Management, R. Hudson (Ed.), Nova
Science Publishers, New York, pp. 41–78, 2015.
[53] D. Simon, A. Shah, C. Scheidegger, ‘Distributed learning with
biogeography-based optimization: Markov modeling and robot con-
trol,” Swarm and Evolutionary Computation, vol. 10, pp. 12–24, June
2013.
[54] M. Hamandi, M. D’Arcy, P. Fazli. ‘Learning to Navigate Like Humans,’
in Proceedings of the Workshop on Learning and Inference in Robotics,
Robotics: Science and Systems Conference (RSS), 2018.
[55] M. Hamandi, P. Fazli, ‘Learning Human Navigational Intentions,’ in
Proceedings of the Workshop on Towards a Framework for Joint
Action: What about Theory of Mind?, Robotics: Science and Systems
Conference (RSS), 2018.
[56] M. Hamandi, M. D’Arcy, P. Fazli, ‘DeepMoTIon: Learning to Navigate
Like Humans,’ in Proceedings of the IEEE International Conference on
Robot and Human Interactive Communication (RO-MAN), 2019.
[57] M. Hamandi, P. Fazli, ‘Online Learning of Human Navigational
Intentions,’ in Proceedings of the International Conference on Social
Robotics (ICSR), pp. 1–10, 2018.
284 Safe Navigation of an Autonomous Robot in Dynamic

[58] S. Karapinar, S. Sariel, ‘Cognitive robots learning failure contexts

through real-world experimentation,’ in Autonomous Robots, 39,
pp. 469–485, 2015.
[59] T. Hoshino, D. Mitsumoto, T. Nagano, ‘Fractal fitness landscape and
loss of robustness in evolutionary robot navigation,’ in Autonomous
Robots 5, pp. 199–213, 1998.
[60] P. Benavidez, M. Jamshidi, ‘Mobile robot navigation and target track-
ing system,’ in 6th International Conference on Systems of Systems
Engineering, Albuquerque, New Mexico, USA, pp. 299–304, 2011.
[61] N.B. Hui, D.K. Pratihar, ‘A comparative study on some navigation
schemes of a real robot tackling moving obstacles,’ in Robotics and
Computer-Integrated Manufacturing, Vol. 25, Issues 4–5, pp. 810–828,
2009.
[62] M.R. Akbarzadeh, K. Kumbla, E. Tunstel, M. Jamshidi, ‘Soft computing
for autonomous robotic systems,’ in Int. J. Computers and Electrical
Engineering, vol. 26, 1, pp. 5–32, 2000.
[63] Y. Kondratenko, E. Gordienko, ‘Implementation of the neural networks
for adaptive control system on FPGA,’ in Annals of DAAAM for 2012 &
Proceeding of the 23rd Int. DAAAM Symp. “Intelligent Manufacturing
and Automation,” Volume 23, No.1, B. Katalinic (Ed.), Published by
DAAAM International, Vienna, Austria, EU, pp. 389–392, 2012.
[64] Y.P. Kondratenko, O.V. Kozlov, O.S. Gerasin, Y.M. Zaporozhets, ‘Syn-
thesis and research of neuro-fuzzy observer of clamping force for mobile
robot automatic control system,’ in IEEE First International Conference
on Data Stream Mining& Processing (DSMP), pp. 90–95, 2016. DOI:
10.1109/DSMP.2016.7583514
[65] O. Gerasin et al., ‘Neural controller for mobile multipurpose cater-
pillar robot,’ in Proceedings of the 2019 10th IEEE International
Conference on Intelligent Data Acquisition and Advanced Computing
Systems: Technology and Applications, IDAACS, pp. 222–227, 2019.
DOI: 10.1109/IDAACS.2019.8924321
[66] D. Simon, ‘The Application of Neural Networks to Optimal Robot
Trajectory Planning,’ Robotics and Autonomous Systems, vol. 11, no.
1, pp. 23–34, May 1993.
[67] C. Churavy, et al., ‘Effective Implementation of a Mapping Swarm of
Robots,’ IEEE Potentials, Vol. 27, no. 4, pp. 28–33, July 2008.
[68] W. Li, C. Ma, F.M. Wahl, ‘A neuro-fuzzy system architecture for
behavior-based control of a mobile robot in unknown environments,’
in Fuzzy Sets and Systems, Vol. 87, Issue 2, pp. 133–140, 1997.
 References 285

[69] A.V. Savkin, C. Wang, ‘Seeking a path through the crowd: Robot naviga-
tion in unknown dynamic environments with moving obstacles based on
an integrated environment representation,’ in Robotics and Autonomous
Systems, Vol. 62, Issue 10, pp. 1568–1580, October 2014.
[70] T.K. Kang, et al., ‘Ego-motion-compensated object recognition using
type-2 fuzzy set for a moving robot’, in Neurocomputing, Vol. 120,
pp. 130–140, 2013.
[71] M. Wang, J.N.K. Liu, ‘Fuzzy logic-based real-time robot navigation in
an unknown environment with dead ends,’ in Robotics and Autonomous
Systems, Vol. 56, Issue 7, pp. 625–643, 2008.
[72] A. Pandey, D.R. Parhi, ‘MATLAB Simulation for mobile robot naviga-
tion with hurdles in cluttered environment using minimum rule based
fuzzy logic controller,’ in Procedia Technology, Vol. 14, pp. 28–34,
2014.
[73] P. Bigai, J. Bartoszek, ‘Low time complexity collision avoidance method
for autonomous mobile robot,’ in Intelligent Systems 2014, Proceed-
ings of the 7th Intern. Conference IS’2014, September 24–26, Warsaw,
Poland, Vol. 2, Springer, Cham – London, pp.141–152, 2014.
[74] P.G. Zavlangas, S.G. Tzafestas, ‘Industrial robot navigation and obstacle
avoidance employing fuzzy logic,’ in Journal of Intelligent and Robotic
Systems, 27(1–2), pp. 85–97, 2000.
[75] D. Simon, H. El-Sherief, ‘Real-Time Navigation Using the Global Posi-
tioning System,’ in IEEE Aerospace and Electronic Systems Magazine,
vol. 10, no. 1, pp. 31–37, 1995.
[76] A.V. Palagin, V. N. Opanasenko, ‘Reconfigurable-computing technol-
ogy,’ in Cybernetics and Systems Analysis 43, no. 5, pp. 675–686,
2007.
[77] Lin, Mingjie, et al., ‘Performance benefits of monolithically stacked 3-D
FPGA’, in Computer-Aided Design of Integrated Circuits and Systems,
IEEE Transactions on 26, no. 2, pp. 216–229, 2007.
[78] A. Drozd, et al. ‘Green Experiments with FPGA,’ in Green IT Engi-
neering: Components, Networks, and Systems Implementation, V.
Kharchenko, et al. (Eds), Vol. 105. Berlin, Heidelberg: Springer Interna-
tional Publishing, pp. 219–239, 2017. DOI: 10.1007/978-3-319-55595-
9_11
[79] A. Drozd, et al. ‘Checkable FPGA Design: Energy Consumption,
Throughput, and Trustworthiness,’ in Green IT Engineering: Social,
Business and Industrial Applications, Studies in Systems, Decision,
and Control, V. Kharchenko, et al. (Eds), Vol. 171. Berlin, Heidelberg:
286 Safe Navigation of an Autonomous Robot in Dynamic

Springer International Publishing, pp. 73–94, 2018, DOI: 10.1007/978-

3-030-00253-4_4.
[80] Y.P. Kondratenko, V.M. Kuntsevich, A.A. Chikrii, V.F. Gubarev (Eds.),
‘Advanced Control Systems: Theory and Applications,’ Series in
Automation, Control and Robotics, River Publishers, Gistrup, 2021.
[81] I. Atamanyuk et al., ‘Generalized method for prediction of the elec-
tronic devices and information systems’ state,’ in 14th International
Conference on Perspective Technologies and Methods in MEMS
Design, MEMSTECH 2018 – Proceedings, Lviv, Ukraine, pp. 18–22,
April 2018. DOI: 10.1109/MEMSTECH.2018.8365709
[82] I. Atamanyuk et al., ‘Method of Polynomial Predictive Control of
Fail-Safe Operation of Technical Systems’, in Proc. XIIIth Int. Conf.
The Experience of Designing and Application of CAD Systems in
Microelectronics, CADSM 2015, 19–23 Febr., 2015, Polyana-Svalyava,
Ukraine, pp. 248–251, 2015. DOI: 10.1109/CADSM.2015.7230848
[83] O. Drozd, M. Kuznietsov, O. Martynyuk, et al., ‘A method of
the hidden faults elimination in FPGA projects for the criti-
cal applications,’ in Proceedings of 2018 IEEE 9th International
Conference on Dependable Systems, Services, and Technologies
(DESSERT’2018), 24–27 May, Kyiv, Ukraine, pp. 231–234, 2018. DOI:
10.1109/DESSERT.2018.8409131J.
[84] I. Atamanyuk et al., ‘Calculation Methods of the Prognostication of
the Computer Systems State under Different Level of Information
Uncertainty,’ in Proceedings of the 12th Int. Conf. on Information and
Communication Technologies in Education, Research, and Industrial
Application. Integration, Harmonization, and Knowledge Transfer, V.
Ermolayev, et al. (Eds), ICTERI’2016, Kyiv, Ukraine, CEUR-WS, Vol-
1614, pp. 292–307, 2016. http://ceur-ws.org/Vol-1614/paper_79.pdf
[85] Y. Kondratenko, P. Khalaf, H. Richter, D. Simon, ’Fuzzy Real-Time
Multi-Objective Optimization of a Prosthesis Test Robot Control Sys-
tem,’ in Y. Kondratenko, A. Chikrii, V. Gubarev, J. Kacprzyk (Eds)
Advanced Control Techniques in Complex Engineering Systems: The-
ory and Applications, Springer, pp. 165–185, 2019.
[86] Y. Kondratenko, G. Khademi, V. Azimi, D. Ebeigbe, M. Abdel-
hady, S.A. Fakoorian, T. Barto, A.Y.Roshanineshat, I. Atamanyuk,
D. Simon, ‘Information, Communication, and Modeling Technolo-
gies in Prosthetic Leg and Robotics Research at Cleveland State
University,’ in Proceedings of the 12th International Conference on
Information and Communication Technologies in Education, Research,
 References 287

and Industrial Application. Integration, Harmonization and Knowl-

edge Transfer, ICTERI’2016, Kyiv, Ukraine, V. Ermolayev, et al.
(Eds), CEUR-WS, Vol-1614, pp. 168–183, 2016. http://ceur-ws.org/Vol-
1614/paper_34.pdf
[87] A. Piegat, ‘Fuzzy Modeling and Control,’ Springer, Heidelberg, 2001.
[88] T. Runkler, ‘Extended defuzzification methods and their properties’,
in Fuzzy Systems, 1996., Proceedings of the Fifth IEEE International
Conference on, Vol. 1, pp. 694–700, 1996.
[89] T. Runkler, ‘Selection of appropriate defuzzification methods using
application specific properties,’ in IEEE Transactions on Fuzzy Systems,
5.1, pp. 72–79, 1997.
[90] J.M. Merigo, A.M. Gil-Lafuente, R.R. Yager, ‘An overview of fuzzy
research with bibliometric indicators,’ in Applied Soft Computing, No.
27, pp. 420–433, 2015.
[91] A. Kaufmann, M. Gupta, ‘Introduction to Fuzzy Arithmetic: Theory and
Applications,’ in Van Nostrand Reinhold Company, New York, 1985.
[92] Y.P. Kondratenko, N.Y. Kondratenko, ‘Soft Computing Analytic Mod-
els for Multiplication of Asymmetrical Fuzzy Numbers,’ in Studies
in Fuzziness and Soft Computing, Volume 393, pp. 201–214, 2021.
DOI: 10.1007/978-3-030-47124-8_17
[93] I.P. Atamanyuk, et al. ‘The Algorithm of Optimal Polynomial Extrapo-
lation of Random Processes.’ In: K.J. Engemann et al. (Eds) Modeling
and Simulation in Engineering, Economics, and Management. MS 2012.
Lecture Notes in Business Information Processing, Vol. 115. Springer,
Berlin, Heidelberg, pp. 78–87, 2012. https://doi.org/10.1007/978-3-642-
30433-0_9
[94] H. Richter et al., ‘Dynamic Modeling, Parameter Estimation and Control
of a Leg Prosthesis Test Robot,’ Applied Mathematical Modelling, Vol.
39, no. 2, pp. 559–573, January 2015.
[95] D. Davis, P. Supriya, ‘Implementation of Fuzzy-Based Robotic Path
Planning.’ In: S. Satapathy et al. (Eds) Proceedings of the Second Inter-
national Conference on Computer and Communication Technologies.
Advances in Intelligent Systems and Computing, vol 380. Springer, New
Delhi, 2016. https://doi.org/10.1007/978-81-322-2523-2_36
[96] T. Jin, B.-J Choi, ‘Obstacle avoidance of mobile robot based on behavior
hierarchy by fuzzy logic.’ Int. J. Fuzzy Logic Intell. Syst. Vol. 12, No.
3, pp. 245–249, 2012.
288 Safe Navigation of an Autonomous Robot in Dynamic

[97] R. Zhao, H.-K. Lee, ‘Fuzzy-based Path Planning for Multiple Mobile
Robots in Unknown Dynamic Environment,’ Journal of Electrical Engi-
neering and Technology, vol. 12, no. 2, pp. 918–925, Mar. 2017.
[98] Z. Gomolka, et al., ‘From homogeneous network to neural nets with
fractional derivative mechanism.’ In: Rutkowski, L. et al. (Eds), Int.
Conf. on Artificial Intelligence and Soft Computing, ICAISC-2017, Part
I, Zakopane, Poland, 11–15 June 2017, LNAI 10245, Springer, Cham,
2017, pp. 52–63. DOI: https://doi.org/10.1007/978-3-319-59063-9_5
[99] G. Khademi, et al. ‘Hybrid invasive weed / biogeography-based opti-
mization,’ Engineering Applications of Artificial Intelligence, Vol. 64,
pp. 213–231, July 2017.
 12
Algorithmic Procedures Synthesis
of Robust-Optimal Control
for Moving Objects

V. L. Timchenko, and D.O. Lebedev

Department of Computerized Control Systems, National Shipbuilding

University, Geroev Ukraine ave. 9, 54025, Mykolayiv, Ukraine
E-mail: [email protected]; [email protected]

Abstract
The transient process control of a variety of moving objects described by
ordinary nonlinear differential equations requires development of robust-
optimal systems with variable structure. Using general algorithmic procedure
for constructing optimal trajectories, determining switching moments and
synthesizing control functions for multidimensional systems are suggested.
The control of the mismatch between the physical object trajectory and
the optimal model calculated trajectory enables to take the values of opti-
mal control into account and to form a robust subsystem, which provides
invariance to incomplete information about the moving object. Simulating
quadrotor UAV stabilization and sea vessel maneuvering under conditions of
external disturbances and parametric noise demonstrates the required control
accuracy in the vicinity of the formed optimal trajectory of the controlled
coordinates. The algorithms and circuit solutions for the synthesis of robust-
optimal variable-structure systems for multidimensional nonlinear moving
objects have been developed as a practical basis of automated procedures
of the synthesis of robust-optimal systems and development of software tools
for various types of moving object control systems.

289
290 Algorithmic Procedures Synthesis of Robust-Optimal Control

Keywords: Algorithmic procedure, robust-optimal control, variable-

structure system, marine moving object, quadrotor UAV.

12.1 Introduction
The original classical works of R. Bellman, R. Kalman, N. Krasovsky, A.
Feldbaum, L. Pontryagin and other well-known scientists demonstrate the
rapid development of the foundations of modern control theory related to the
problems of analyzing desired (optimal) trajectories of object movement.
The basic requirements for the tasks of the optimal control synthesis
for quadratic criteria of optimality and maximum speed were mathemati-
cally formalized, supported by algorithms for the determination of switching
moments. It was a breakthrough in control systems, especially in the space
sector, where the highest accuracy and efficiency were required for well-
defined objects and given properties of the environment [1, 2]. It is well
known there are many control problems, e.g. motion under conditions of
aerodynamic and/or hydrodynamic resistance of a stochastic medium with
incompletely known parameters. However, the deviation of a mathematical
model from a moving physical object had so significant impact on control
errors that it stimulated the rapid development of methods of adaptive, and
later, robust control. Later, summing up the development of these methods,
V. Kuntsevich [3] classified them as adaptive-optimal methods requiring
identification of parameters and adjustment of control (which is rather dif-
ficult to provide in real-time), and robust-optimal methods that create control
algorithms with known stable solutions concerning bounded uncertainties.
These methods were further developed in subsequent works [4–8].
Other widely used methods based on robust properties of sliding modes
(considering the additional power of control for their implementation), but
frequent switching of control devices negatively effects their efficiency
[8–11]. Robust properties are also inherent in PID controllers, since they use
approximate tuning for the parameters of the plant model, respond to an error,
and compensate uncertain disturbances [12].
The complexity of the synthesis of multidimensional objects control was
revealed and caused the development of a new approach for forming the
desired trajectories in solving inverse problems of dynamics. The symmetry
of solutions of the motion modeling and control synthesis made it possible
to focus on developing the programmed trajectories of moving objects at
the initial stage of control synthesis. These methods simplify the synthesis
of optimal control since they do not require solving Riccati equations or
 12.1 Introduction 291

variational problems. When forming the corresponding requirements for the

controlled process and the desired (optimal) trajectories of the moving object,
the control is based on solving the inverse problems of dynamics with high
dynamic accuracy ensuring the properties of robustness [13–18]. The prob-
lems of synthesis of optimal stabilization systems for the output coordinates
of various moving objects, including additional qualities of the transient
process, are solved on the basis of Lyapunov matrix algebraic equations for
the quadratic criterion by the synthesis of optimal controllers in static and
dynamic feedbacks even during stabilization of various objects [19–21].
In V. Larin’s [22], the important problem of determining the coefficients
in the weight matrices of a quadratic performance criterion is solved as
an inverse problem for the synthesis of an optimal controller of a linear
stationary object, i.e. inversion of the problem of “analytical design of optimal
controllers”. It should be noted that the computational complexity of the
analytical design of suboptimal controllers increases significantly with an
increase in the order of the controlled object and does not allow their practical
use in the synthesis of control systems [14]. The use of the principles of modal
control and Ackerman formula presupposes on heuristic assumptions about
the optimality of the transient process and allows the formation of the desired
trajectories for various distributions of parameters to ensure the appropriate
type of the transient process. During the synthesis of scalar control, the corre-
sponding additional matrices of parameters are rather conditionally assigned.
The methods of high computational efficiency based on H optimization have
also been developed, primarily for high order moving objects limited mainly
by linear systems [23–24].
Some particular characteristics of modern control methods for several
classes of moving objects such as marine moving objects and unmanned
aerial vehicles, which have been widely applied in modern transport, military
and other similar infrastructural systems, are considered in detail. The models
of the dynamics of a marine moving object as a rigid body functioning at
the interface between air and liquid media have been sufficiently studied
[25–28] and are described by multidimensional nonlinear stochastic differ-
ential systems of equations. A comprehensive and accurate description of the
dynamics of a marine moving object under the influence of random distur-
bances (currents, wind, and sea waves) seems extremely difficult due to the
incomplete definiteness of the dynamic model parameters, as it is impossible
to physically form and mathematically describe all the factors that effect the
value of parameters or their function and also to measure some components of
external disturbances (wind pulsations, irregular waves). The main applicable
292 Algorithmic Procedures Synthesis of Robust-Optimal Control

simplifications for the considered problems of marine moving objects control

include of external random stationary disturbances, the horizontal plane of
dynamic vibrations of the object, and several other assumptions. A high level
of computational complexity limits the use of classical methods of optimal
control, considering the multidimensional nonlinear model of the controlled
object, the solution of boundary value problems for the synthesis of program
control, and the solution of Riccati matrix equations when designing an opti-
mal controller. The criteria of quadratic optimality require the formalization
of weighting coefficient matrices based on additional calculation procedures
[22] or empirical calculations.
The development of the fundamental feedback principle of control [8]
is based on the wide application of static feedbacks for the synthesis of
optimal control [21], linear matrix inequalities taking into account the action
of arbitrary, bounded perturbations [19], however, they are mainly used for
linear or equivalent linear models of controlled systems, incl. marine moving
objects [29, 30]. Some alternative approaches, e.g. based on the principles of
solving inverse problems of dynamics and structural synthesis [13, 15], do
not always enable to optimize the parameters of programmed (assigned) tra-
jectories. The incomplete definiteness of the parameters of both the dynamic
model of marine moving object and the environment, as well as the problem
of real-time control, leads to the necessity to apply the principles of robust
control. For example, the use of variable structure systems [10, 11], that is
effective for a certain (reactive) type of propellers, requires the increased
control energy consumption and frequent switching, which can lead to a
decrease in operability for the controls of a marine moving object.
The implementation of high-tech functional requirements for maneu-
vering and positioning of marine moving objects can be carried out using
applied control systems with a variable structure based on special switching
feedbacks. These systems ensure the minimization of energy consumption
or maximization of performance for a specific functional task of marine
mobile object control (respectively, working technological or transient critical
modes) with the required control accuracy and sufficient invariance to the
uncertainty of object parameters and the environment [31–36].
The deployment of unmanned aerial vehicles (UAV) enables to signif-
icantly expand the use of visual operational monitoring tools of technical
objects for various purposes. For example, safe navigation is ensured by the
increasing safety requirements for restricted water areas, especially during
various technological operations, such as maneuvering, towing, bunkering,
loading etc. The analysis of the marine accidents which happened in restricted
 12.1 Introduction 293

water areas proves that about 70% of accidents occur due to human factor. It
is partly due to the lack of sufficient traffic safety data for operators to make
decisions in extreme operational conditions.
The use of UAV (quadrotor UAV) for marine environment monitoring
makes it possible to expand the navigation safety information assessment
and to increase the control of the marine moving objects improving in
this way the marine emergency prevention. It is necessary to improve the
quality control systems to ensure the optimal trajectory of stabilization of
the quadrotor UAV along all given controlled coordinates (the dynamics are
described by a system of six differential equations of the second-order [37,
38]). Quadrotor UAV functions under conditions of limited rotors power and
incomplete information of uncontrolled external disturbances (the complexity
of a complete description of the physical impact of wind perturbation on a
UAV involves the selection of the following components of airflow: high-
frequency random pulsations; low-frequency piecewise constant components
in horizontal (constant wind) and vertical directions [39]).
The LQR method is widely used for dynamic system stabilization as a
modern and effective method to synthesize optimal controllers for UAVs at
a minimal energy cost [40–45]. The comparison between LQR and the PID
algorithm shows that LQR has better energy efficiency than PID. At the same
time, LQR algorithms working well in conditions of noisy control processes
do not ensure accurate control in conditions of uncontrolled disturbance
compared to the control systems with the sliding modes [46–48]. The use
of control synthesis based on quadratic functionals and Lyapunov functions
[49–51] is limited to linear systems. Accordingly, the main problem of the
QFT method consist in the formation of feedback under the conditions of
uncertainty of disturbance parameters [52–54]. H 2 /H control synthesis is
increasingly used for the synthesis of robust controllers. This optimization-
based method is associated with the mathematical description of the expected
behavior of a closed system and its stability [55–60]. It has a strictly math-
ematical basis for the optimization process thus having been applied to both
classical optimal and robust control. There are some advantages of the linear
approach due to its powerful tools for the design and analysis of controllers.
This method considers the resistance to incoming perturbations (aerodynamic
moments and forces of wind) as the main source of errors in orientation
[61–64]. Neural networks are used to synthesize both robust and optimal
controllers [65, 66]. Wind disturbances are compensated by an artificial
neural network, the parameters of which are configured in the current mode of
operation. The controller is complemented by multilayer neural networks and
294 Algorithmic Procedures Synthesis of Robust-Optimal Control

adaptive control to compensate uncontrolled disturbances. A large volume of

online computing required reduces the effectiveness of this approach. The
above problems are related to the multidimensionality and nonlinearity of the
control object, as well as the uncertainty of wind disturbances significantly
complicates the process of synthesis and real-time control.
The proposed approach for the synthesis of robust-optimal systems is
based on the use of the system with variable feedbacks and is described for
models of marine moving objects [31–36] and quadrotor UAV [67–70]. It
is possible to use this system for multidimensional nonlinear non-stationary
models of moving objects. The other advantage of this method includes the
control synthesis using direct optimality conditions based on energy analysis
and robust correction of control processes based on minimizing control
errors.

12.2 Problem Review

It is possible to determine how to optimally control dynamic systems under
uncertainty conditions for mathematical models of objects and the environ-
ment by designing efficient automatic control systems in terms of quality
indicators. The criteria of optimality for control processes are primarily deter-
mined by the necessity to reduce energy costs under the conditions of limited
control resources or maximum speed of transient processes at critical modes
of the control object operation. The criteria of optimality that mathematically
determine these requirements are opposite in terms of minimax estimates.
However, if the quality indicators of specific stages of technological processes
are considered as the basis for optimization, then it would be possible to
distinguish the stage of the functioning of a dynamic system at the energy
consumption minimum and the stage of critical control with the maximum
speed of transient processes. Thus, two separate stages of optimization are
determined making it possible to apply the corresponding optimality criteria
and the procedure for the optimal control synthesis.
The aim of the research is designing a generalized approach to the
synthesis of high-precision and robust-optimal control systems that are uncer-
tainties stable and applicable for multidimensional nonlinear moving object.
The control of dynamic systems, which include moving objects under the
influence of external and parametric disturbances, are considered at transient
modes, i.e. generally at stabilization modes of controlled coordinates. The
mathematical description (simulation) of dynamic systems in the form of
ordinary differential equations represents the structural-parametric model of
 12.2 Problem Review 295

the control object. The structure of the interaction between the object and the
environment is determined by the type of differential equations and is suffi-
ciently definite. The structure can be changed due to the correct simplification
of the system of equations, e.g. by replacing the nonlinear model with a linear
model. However, the values of the parameters (coefficients) of the model
of the controlled object are rather undefined values, and they describe an
incompletely informative model. While designing automatic control systems
for dynamic objects, the parameters of the mathematical model of the object,
which are the basis of synthesis, differ from the calculated ones in real
systems due to the impossibility to physically determine and mathematically
describe all factors that effect the value or function of the parameter. For
example, hydrodynamic drag coefficient and the attached mass of the vessel
depend, among other things, on the shape of the vessel hull, which can be
changed during operation (including overgrowing with marine organisms),
and the density of the marine environment, the value of which is also not
constant for different areas of operation of the vessel.
The disturbances under which the controlled object functions can be
divided into controlled (measured, physically measurable) and uncontrolled
(unmeasured), e.g. wind speed or wind direction that can be accurately
measured using an anemometric system. It should be noted that if the control
object does not allow the placement of the anemometric system on the
hull, e.g. a quadrotor UAV, then, in this case, the wind effect is uncontrol-
lable. Irregular sea waves that determine vessel pitching are uncontrollable
disturbances.
The system of equations for the generalized model of the nonlinear non-
stationary dynamic system of a moving object is described by Newtonian
mechanics in vector-matrix form

Ẍ(t) = AX Ẋ(t) + BX U(t) − g + CX F(t). (12.1)

For optimal control problems, the requirements are formed to ensure

the optimal trajectory Xopt (t) of stabilization of the dynamic system along
all given controlled coordinates considering the fulfillment of the optimal
criterion for the minimum control energy consumption U(t)
 t2
J= Q (X, U) dt = min. (12.2)
t1

The maximum speed of transient processes in dynamic systems is

described by ensuring the minimum time interval t1 ÷t2
296 Algorithmic Procedures Synthesis of Robust-Optimal Control
 t2
J= dt = min. (12.3)
t1

Thus, the task is to develop practical algorithmic procedures for construct-

ing multidimensional object optimal phase trajectories based on the existing
variety of trajectories connecting the initial and final points of the space R,
satisfying the control constraints and additional requirements–the minimum
energy consumption or the minimum time of the transient process (maximum
speed).
Suppose that the planning of the trajectory of a nonlinear non-stationary
moving object can be described in the form of certain segments with constant
values of the corresponding ith (or, respectively, zero (i + 1) - th) derivatives of
the state coordinates X(t) and, thus, described by polynomial dependencies.
The type of polynomial dependencies and the required number of trajectory
segments are determined by the type of specified optimality criteria and the
values of the boundary conditions.
The transition of a moving object from the initial segment of the trajectory
to a given segment, considering the control constraints, can be described by
the following equations for the state coordinate vector
(m) (tsi − tsi−1 )m
Xopt (tsi ) = X(tsi−1 ) + ... ± X (tsi−1 ) . (12.4)
m!
The solution to the general problem of the dynamic object control must
consider the uncertainty of the mathematical model of the object and external
perturbations using robust control methods. The robust-optimal approach
considered involves the synthesis of a robust correction loop that compensates
deviations from the optimal trajectory of the moving object with minimal
vector errors E(t) and its derivatives

Ë(t) + G1 Ė(t) + G2 E(t) → min ≈ 0. (12.5)

12.3 Optimal Synthesis

The approach to the synthesis of the optimal system is based on the system
with a variable feedback structure. This approach includes the following main
steps:
1. Optimal trajectory planning.
2. Determining switching moments of control functions in the feedback
loops.
 12.3 Optimal Synthesis 297

3. Synthesizing optimal control functions in the corresponding feedback

loops.
Considering a priori known parameters of the model of a dynamic object, the
problem is mathematically exactly solved as an optimal control problem.

12.3.1 Optimal Trajectory Planning

Optimal phase trajectories of the moving object stabilization can be formed
using polynomial or exponential representations. Formalizing the trajectory
type is determined by the requirements for the optimality of the control
process, its type, and values of the boundary conditions, as well as by the
constraints on the control actions.
Considering the optimality criterion, one can be guided by the following
functional requirements: minimum energy consumption in operating and
technological modes of moving object control (taking into account energy
consumption for stabilizing the moving object at a given altitude) and
maximum speed at critical, extreme modes of operations.
For a multidimensional object of dimension j = 1,. . . ,n with the output
vector of phase coordinates X(t) boundary conditions are defined as some
(i) (i)
points in the space of variables R for initial X(0) and final X(T ) bound-
ary conditions. The requirement for physical feasibility of control actions
imposes restrictions on their maximum values, and for the polynomial form
of trajectories defined by certain dependencies between the values of controls
and derivatives of the output coordinates.
When assessing the control energy costs, the optimality criterion is
applied, which physically expresses the average control power in the follow-
ing functional (12.2) in the expanded quadratic form for control power
 T
J= UT QUdt = min . (12.6)
0

Thus, the task is to form the algorithmic procedures for constructing

optimal phase trajectories that are practically realizable for multidimensional
objects based on the existing variety of trajectories connecting the initial and
final points of the space R, to satisfy the control constraints and additional
requirements–the minimum energy consumption or the minimum time of the
transient process (maximum speed).
To simplify the presentation (without loss of generality), consider a
linear moving object’s model in the horizontal plane without considering
298 Algorithmic Procedures Synthesis of Robust-Optimal Control

perturbations
Ẍ(t) = AẊ(t) + BU(t). (12.7)
Work during the movement of the object (in the absence of external
perturbations) along the trajectories of stabilization is performed by control
and damping (aerohydrodynamic resistance) forces. To estimate the energy
losses due to damping forces, the Rayleigh dissipative function must be
introduced  T
T
D(Ẋ) = −0.5 Ẋ AẊdt. (12.8)
0
To optimize the problem of minimum energy consumption, lets consider
the Equations (12.7, 12.8), the minimized functional in the form of Lagrange
function can be written as
T
L = −0.5Ẋ (t)AẊ(t) + 0.5UT (t)QU(t) + λT [Ẍ(t) − AẊ(t) − BU(t)].
(12.9)
The extremum of functional (12.9) is determined by the following Euler–
Lagrange equation considering Ẋ(t) = V(t)
∂L d ∂L ∂L d ∂L
− = − = −AT Ẋ(t) − AT λ(t) − λ̇(t) = 0,
∂V(t) dt ∂ V̇(t) ∂ Ẋ(t) dt ∂ Ẍ(t)
(12.10)
and for control
∂L
= QU(t) − BT λ(t) = 0. (12.11)
∂U(t)
From the Equation (12.11), optimal control can be expressed as
Uopt (t) = Q−1 BT λ(t). (12.12)
When solving the Equation (12.10), it is assumed that polynomial rep-
resentation of the moving object trajectories is used. In this case, the object
can be at rest or perform a uniform (steady-state relative to the speed Ẋ(t)
with zero acceleration), or uniformly accelerated (steady-state relative to
acceleration Ẍ(t) with zero third derivative), or established for the third
...
derivative X (t) (with zero fourth derivative) and further, established for mth
(m)
derivative X (t) (with zero (m+1)-th derivative), motion.
(m) (m)
In general, for X (t) = X (0) = const the equation of one segment of
the trajectory for phase coordinate vector will take the form
! (i)
m−1
ti (m) tm
X(t) = X(0) ± X (0) . (12.13)
i! m!
i=0
 12.3 Optimal Synthesis 299

Assuming that to achieve a given value we need two segments (accel-

eration and deceleration), the equations of optimal trajectories will take the
following form for joint switching moments ts1 and T for all state coordinates
of a moving object
1st segment 2nd segment
⎧
⎪ 
m−1 (i) (ts
1)
i (m) ⎧
⎪
⎪X(t1 ) =
s
X (0) i! + X (0) ⎪  (i) s (T −ts1 )i
m−1 (m) (T −ts m
⎪ ⎪ 1)
⎪
⎪ i=0 ⎪
⎪X(T ) = X (t1 ) − X (0) ;
⎪
⎪ ⎪
⎪
i! m!
⎪
⎪ (ts m ⎪
⎪
i=0
⎪ 1) ⎪
⎪
⎪ ; ⎪
⎪
⎪
⎨
m! ⎪
⎨  (i+1) s (T −ts1 )i
m−1 (m) (T −ts
1)
m−1
 (i+1)
m−1
(ts
1)
i Ẋ(T ) = X (t1 ) − X (0) (m−1)! ;
⎪
⎪Ẋ(ts1 ) = X (0) i! ; ⎪
⎪
i!
⎪
⎪ ⎪
⎪
i=0
⎪
⎪ i=0 ⎪
⎪.......
⎪
⎪....... ⎪
⎪
⎪
⎪ ⎪
⎪
⎪
⎪ ⎪
⎪
⎪
⎪ ⎩(m−1) (m−1) (m)
⎩(m−1) s (m−1) (m) X (T ) = X (ts1 ) − X (0)(T − ts1 ).
X (t1 ) = X (0) + X (0)ts1 .
(12.14)
The Equation (12.10) can be written for the trajectory with a transient time T
with a zero third derivative (m = 2) in the form

λ̇(t) + AT λt − AT (Ẋ0 + Ẍ0 T ) = 0 (12.15)

Together with the equation for the controlled system and control (12.12)

Ẍ(t) = AẊ(t) = BQ−1 BT λ(t) (12.16)

Equation (12.15) forms a boundary value problem for optimal control, the
solution of which for a trajectory with zero third and fourth derivative can
also be described in the polynomial form

U(t) = U(tsi ) + U̇(tsi )tsi ,

(tsi )2
U(t) = U(tsi ) + U̇(tsi )tsi + Ü(tsi ) .
2
In general form, for the mth derivative we obtain control in the form

! (i)
m−2 s
s t
U(t) = U(ti ) i , (12.17)
i!
i=0

(i)
where order m and values U(tsi ) are limited by the physical constraints of the
controls.
For the polynomial representation of trajectories, each coordinate is inde-
pendent of the others (although in the general case considered below, they
300 Algorithmic Procedures Synthesis of Robust-Optimal Control

have different end times of the transient process). It should also be noted
that for a zero matrix A, the polynomial and exponential representations
of the trajectories coincide. Thus, the polynomial representation of the tra-
jectories, in contrast to the exponential one, makes it possible to construct
the moving object stabilization trajectories in advance (before the synthesis
of the control) rather simply. The polynomial representation of trajectories
is applicable for a complete mathematical description of the motion of the
moving object, i.e. considering nonlinearity and non-stationarity. With the
exponential representation of the trajectories, it is necessary to numerically
solve nonlinear non-stationary differential equations of a higher order which
describe the dynamics of the object.
The use of the polynomial form of trajectories makes it possible to
estimate the dependence of the optimality functional for the minimum energy
consumption on the time of the transient process. Consider the motion of the
moving object along the trajectory segments (12.4) for the 3rd and 4th zero
derivatives for the coordinate vector X(t), as well as the final conditions (for
simplicity, X0 = Ẋ0 = 0)
...
X(t) = 0.5Ẍ0 t2 ; Ẍ(t) = Ẍ0 = const; X (t) = 0; X(T ) = Xk ;
(12.18)
1 1 ... 1 ... ...
X(t) = Ẍ0 t2 + X0 t3 ; Ẋ(t) = Ẍ0 t + X0 t ; Ẍ(t) = Ẍ0 + X0 t;
2
2 6 2
... ... (IV )
X (t) =X0 = const; X (t) = 0; X(T ) = Xk ; Ẋ(T ) = Ẋk .
(12.19)
Then it is possible to obtain the necessary initial conditions, respectively,
for the 1st
Ẍ0 = 2T −2 Xk ,
and for the 2nd segment
...
Ẍ0 = T −2 (6Xk − 2Ẋk T ); X0 = T
−3
(6Ẋk T − 12Xk ).
Further, for the 1st segment, we obtain the equation for the initial conditions
for the control
Ẍ0 = AẊ0 + BU0 ⇒ U0 = B−1 Ẍ0 ;
...
−1
X0 = AẌ0 + BU̇0 = 0 ⇒ U̇0 = −B AẌ0 ,
and the general equation for the control in the form of
 12.3 Optimal Synthesis 301

U(t) = U0 + U̇0 t = 2B−1 T −2 Xk − 2B−1 AT −2 Xk t =

= 2B−1 T −2 Xk (E − At). (12.20)
When carrying out similar transformations for the 2nd segment of the
(IV ) ...
trajectory, it is possible to obtain (considering X0 = A X0 +BÜ0 = 0)
U(t) = U0 + U̇0 t + 0.5Ü0 t2 = B−1 T −2 {(6Xk − 2Ẋk T )+
+ [T −1 (6Ẋk T − 12Xk ) − A(6Xk − 2Ẋk T )]t−
− AT −1 (3Ẋk T − 6Xk )}t2 . (12.21)
It should be noted that the equations for the boundary condition Ẋ(T )T
with an increase in the time of the transient process T, as follows from the
analysis of the equations for the jth component of the vector X(t), changes
within narrow boundaries, and for small ẍj0 and sufficiently big Tj , almost
equal to a constant
1 ... 1 ...
2xjk = (ẍj0 + x j0 Tj )Tj 2 ; ẋjk = (ẍj0 + x j0 Tj )Tj ;
3 2
...
2xjk ẍj0 + 13 x j0 Tj 2
= ... ≈ = C.
ẋjk Tj ẍj0 + 12 x j0 Tj 3
Then, for the elements of the functional (12.6) concerning the time of the
transient process T after integration of the form (12.20, 12.21), we obtain the
equations proportional to ∼1/T, which signalizes the decrease in the required
control power during movement of a moving object along the trajectories
of the form (12.3.1), while the time T increases. Thus, we can draw a
generalized conclusion that for a given boundary constraint for the coordinate
vector, the trajectory with the longest possible transient time is optimal in
terms of energy consumption, while the most energy-intensive control is the
trajectory with the maximum possible response time considering the physical
constraints on the values of the maximum control. Trajectories of the form
(12.3.1) of the control object in polynomial form with positive values of the
derivatives of the state coordinate vector is optimal relative to the criterion
(12.6) when a moving object moves with the minimum possible number of
minimum possible values of the derivatives of the coordinate vector
 
(i)
minmin{Ẋ X (t0 ) , t },
i (i)
X (t0 )
fulfilling all specified boundary conditions.
302 Algorithmic Procedures Synthesis of Robust-Optimal Control

For the tasks of maximum speed, the trajectories are optimal relative to
the criterion (12.3) when a moving object moves with the maximum possible
number of maximum possible values of the derivatives of the coordinate
vector, taking into account the constraints on the control action
 
(i)
maxmax{Ẋ X (t0 ) , t },
i (i)
X (t0 )

(i)
and the fulfillment of all specified boundary conditions X(t0 ) =
(i−1)
f [X(t0 ), U max (t0 )].

12.3.2 Determining Switching Moments of Control Functions

in the Feedback Loops
The introduction of constraints on the control action limits the number of
possible controlled coordinate derivatives, which effects the form of the opti-
mal trajectory and significantly complicates its calculation. The algorithms
of multidimensional system have been developed for the given boundary
conditions and the values of the derivatives of the object coordinates vector
considering the control constraints. These algorithms are based on the solu-
tion of the systems of algebraic equations in the form (12.3.1) and include
the formation of leading, sub-leading and driven variables, the formation of
some sequence of switching moments of control functions in feedbacks of a
controlled object.
To construct an optimal trajectory in terms of the maximum speed (12.3)
and determine the switching moments considering the constraints on the
control action, the algorithm is formed (Figure 12.1) for the 6th order system.
The switching moments for the problem of minimum energy consumption
(12.6) are determined on the basis of the system of algebraic equations in
the form (12.3.1) for two or more segments by calculating the roots that
represent the switching moments tsij and Tj that in general case will differ
for each jth state coordinates. For this, the corresponding algorithm is applied
(Figure 12.2).

12.3.3 Synthesizing Optimal Control Functions in the

Corresponding Feedback Loops
When constructing the optimal model of the control system, we assume that
the dynamics of a moving object is described by the vector-matrix equation
 12.3 Optimal Synthesis 303

Figure 12.1 Algorithmic procedure A for constructing the optimal trajectory in terms of
maximum speed

in general form (12.1), the parameters and external disturbances are precisely
known and specified.
The equations (12.17) define the representation of control functions in
the form of time polynomials, which are switched by certain algorithmic
procedures A and B, and is described by the block diagram shown in
Figure 12.3.
An alternative is to represent controls in the form of integrating func-
tions in the feedback loops of a moving object (Figure 12.4). Therefore,
we use the differential transformation of the vector-matrix Equation (12.1)
considering the above assumptions regarding disturbances, for example, the
third and fourth zero derivatives of the coordinate vector. This differential
304 Algorithmic Procedures Synthesis of Robust-Optimal Control

Figure 12.2 Algorithmic procedure B for the problem of minimum energy consumption

Figure 12.3 Block diagram of the implementation of control in the form of time polynomials
(F(t) is not considered; the index X for the matrices A, B is omitted)

transformation is correct due to a clear indication of the initial conditions of

coordinates and control on each segment of the optimal trajectory. Thus, we
form the corresponding equations for the balance of forces and moments, as
well as their derivatives,
...
for a segment of the trajectory with the fulfillment
of the condition X (t) = 0

AX Ẍ(t) + ȦX Ẋ(t) + ḂX U(t) + BX U̇(t) + ĊX F(t) + CX Ḟ(t) = 0,

(12.22)
 12.4 Robust Correction 305

(IV)
and for the condition X (t) = 0
...
AX X (t) + 2ȦX Ẍ(t) + ÄX Ẋ(t) + BX Ü(t) + 2ḂX U̇(t)+
+B̈X U(t) + CX F̈(t) + ĊX Ḟ(t) + C̈X F(t) = 0. (12.23)

Vector-matrix transformations of the Equations (12.22, 12.23) make it

possible to write them in a form that determines the control vector and ensures
the movement of a moving object along optimal trajectories, respectively, for
the conditions of the third and fourth zero derivatives of the coordinate vector

BX U̇(t) + [AX BX + ḂX ]U(t) =

= −[(A2X + ȦX )Ẋ(t) + (AX CX + ĊX )F(t)+
+ CX Ḟ(t) − AX g]; (12.24)
BX Ü(t) + (AX BX + 2ḂX )U̇(t) + (A2X BX + 2ȦX BX + AX ḂX +
+ B̈X )U(t) = −[(A3X + 2ȦX AX + AX ȦX + ÄX )Ẋ(t)+
+ (A2X CX + 2ȦX CX + AX ĊX + C̈X )F(t)+
+ (AX CX + 2ĊX )F(t) + CX F̈(t) − (2ȦX + A2X )g]. (12.25)

For the proposed representations of the control system, the necessary

initial conditions of the control functions that ensure the movement of the
moving object among the corresponding section of the trajectory are obtained
from the following algebraic equations

U(tsi ) = B−1
X [Ẍ(ti ) − AX Ẋ(ti ) + g − CX F(ti )],
s s s
...
U̇(tsi ) = B−1
X [X (ti ) − AX Ẍ(ti ) − ȦX Ẋ(ti ) − ḂX U(ti ) − ĊX F(ti )+
s s s s s

+ CX Ḟ(tsi )]. (12.26)

It should be noted that the system is shown in Figure 12.3. is simpler to

implement, but the system in Figure 12.4, is implemented by integrating
blocks in feedback loops and has additional filtering properties.

12.4 Robust Correction

The solution to the moving objects control problem under conditions of
parameter’s uncertainty, measurement noise, and external disturbances is
based on the use of a variable structure system forming the optimal reference
306 Algorithmic Procedures Synthesis of Robust-Optimal Control

Figure 12.4 Block diagram of the implementation of control in the form of integrating
functions in the feedback

model of the moving object motion and the corrective control signal formed
by a robust loop based on the mismatch between the optimal and physical
signal output. The integral signal of optimal and robust control is transferred
to a moving object propellers. The general constraint on the control action
should limit the values of the optimal and corrective control
|Uopt | + |Ucor | ≤ |Umax | .
The proposed approach for an admissible simplification of the synthesis of
robust control includes the linearization of the dynamics equations in the
vicinity of the nominal values of the parameters, which makes it possible
to apply the superposition principle to analyze control errors. The optimal
control and the specified trajectories of the moving object are formed, as
described above, taking into account the fact that the nonlinearity of the
model, and the discrepancy arising from the application of linearization can
be attributed to additional uncertainty, which is compensated by a robust
control.
The external disturbances F, some of which Fcont can be measured and
controlled, the parametric measurement noise η, the mismatch between the
parameters of the physical moving object, described by the matrix transfer
 12.4 Robust Correction 307

Figure 12.5 The structure of a linear and stationary robust-optimal system

function Wo (p), and the optimal mathematical model Wopt (p) are considered
to be uncertainties that are compensated by the robust loop.
Based on the structural diagram of a robust-optimal system (Figure 12.5),
the error vector can be determined as
E(p) = [I + Wo (p)Wcor (p)]−1 {[Wopt (p) − Wo (p)]Uopt (p)+
+ Wopt (p)Fcont (p) − Wo (p)F(p) − η(p)} = EU (p)+
+ EF (p) + Eη (p). (12.27)
In a more simplified form, the error vector can be represented in an
approximate form due to the unmeasurable part of the perturbations ΔF(p)
EF (p) ≈ [I + Wo (p)Wcor (p)]−1 Wopt (p)ΔF(p).
The requirements for the structure and parameters of a robust correction in the
form of the minimum H 2 norm [71] of the matrices of sensitivity functions
can be written for:
• input programmed optimal control based on the mismatch between the
parameters of the optimal model and the physical object
SU = [I + Wo (jω)Wcor (jω)]−1 [Wopt (jω) − Wo (jω)];
• the difference between the values of the vector of external disturbances
F(t) and their controlled part Fcont
SF = [I + Wo (jω)Wcor (jω)]−1 Wopt (jω);
308 Algorithmic Procedures Synthesis of Robust-Optimal Control

• values of the parametric noise vector η

Sη = [I + Wo (jω)Wcor (jω)]−1 .
The robustness of the control system is limited when the condition for the
H 2 norm of the sensitivity matrix is met
 
[I + Wo (jω)Wcor (jω)]−1 [Wopt (jω) − Wo (jω)] ≤ 1;
 
[I + Wo (jω)Wcor (jω)]−1 Wopt (jω) ≤ 1; (12.28)
 
[I + Wo (jω)Wcor (jω)]−1  ≤ 1.
The analysis of the equations obtained (12.28) shows that together with the
small mismatch between model and the physical moving object’s matrix norm
Wopt (jω) − Wo (jω) ≤ 1 amplification is applied in the straight part of
the robust loop, which does not have a negative effect on stability but instead
increases the accuracy of the control system.
The structure and parameters of a robust loop can be obtained from
the condition of the minimization of the error vector and its derivatives
in the form (12.5). The equation of the physical moving object (12.1)
(considering only the controlled disturbance Fcont and the use of a robust
loop) takes the following nonstationary form (considering linearized matrices
Ā(t), B̄(t), C̄(t))
Ẍ(t) = Ā(t)Ẋ(t) + B̄(t)[Uopt (t) + Ucor (t)] + C̄(t)Fcont (t) − g. (12.29)
The Equation (12.1) for the optimal moving object model can be
written as
Ẍm (t) = Ām (t)Ẋm (t) + B̄m (t)Uopt (t) + C̄m (t)Fm (t) − g. (12.30)
To determine the control signal using The Equations (12.29) and (12.30),
the approximate equation for the vector of control errors can be written in the
form
Ë(t) ≈ −G1 Ė(t) − G2 E(t) ≈ Ām (t)Ė(t) − B̄m (t)Ucor (t). (12.31)
Using (12.31), the dependence for the robust correction loop vector can
be obtained
Ucor (t) = B̄−1
m (t)[Ām (t)Ė(t) + G1 Ām (t) + G2 E(t)]. (12.32)
The structural implementation of a robust-optimal control system of
variable structure using robust correction is shown in Figure 12.6.
 12.5 Experiments 309

Figure 12.6 The structural implementation of a robust-optimal control system of variable

structure

12.5 Experiments
12.5.1 Sea Vessel Maneuvering
To simulate the process of stabilizing the course angle ϕ (t) = x1 (t) and
angular speed ω (t) = x2 (t) of the vessel, controlled by changing the angle
of the rudder blade α(t), considering the values of the reduced aerohydro-
dynamic coefficients and external disturbance f (t) [27,31], let’s consider the
310 Algorithmic Procedures Synthesis of Robust-Optimal Control

Figure 12.7 Course angle over time

problem of sea vessel maneuvering at maximum speed (12.3)

Ẋ (t) = AX X (t) + BU (t) + Cf (t) ,

⎛ = (x1 (t) x2 (t) x3 (t)) ; U(t)

where X(t) T
⎛ α̇(t)) ;
⎞ = (α(t)
T
⎞
0 1 0 0 0
AX = ⎝ 0 0 1 ⎠; B = ⎝ 0 0 ⎠;
0 −0.03 |x2 (t)| −0.084 0.0002 0.0063
T
C = (0 0 0.00034) .
Simulating the process of vessel stabilization is considered taking into
account the impact of uncontrolled irregular waves [27] Wf (p) and the
mismatch of the parameters of the mathematical model and the physical
vessel, set within 20%, as well as the influence of parametric noise of a given
intensity.
The simulation graphs include time trajectories for course angle ϕ(t)
(Figure 12.7), rudder angle α(t) (Figure 12.8) and errors εϕ (t) along the
course angle (Figure 12.9).
The simulation of the vessel stabilization under the influence of irreg-
ular sea waves, set by the shaping filter for a given wave intensity of four
points, was considered taking into account the noise of measuring the output
coordinates. The specified maneuvering trajectory includes three sections at
different course angles, which is typical when sailing in limited water areas
(sea channels, straits, etc.), as well as, for example when laying underwater
pipelines. The simulation results demonstrate that the control accuracy is
ensured with errors of less than 1%.
 12.5 Experiments 311

Figure 12.8 Rudder angle over time

Figure 12.9 Course angle error over time

12.5.2 Quadrotor UAV Stabilization

The motion dynamics of a quadrotor UAV are described by the system of
six second-order differential Equation (12.1) considering the effects of wind
disturbances. The complexity of a complete description of the physical effect
of wind disturbance on a quadrotor UAV suggests [39] the selection of the
following components of the airflow: high-frequency random pulsations;
low-frequency piecewise constant components in horizontal (constant wind)
and vertical directions (ascending or descending flows). Thus, the quadrotor
UAV motion is described by the following system of differential equations
[37, 41, 68]

Ẋ (t) = A (X) X (t) + B (X) U (t) + CF (t) − g,

312 Algorithmic Procedures Synthesis of Robust-Optimal Control

where X(t) = (vx vy vz ωθ ωψ ωϕ )T ; (F(t) = fx fy fz fθ fψ fϕ )T ;

) *T
g = 0 0 9.81 0 0 0 ;
A (X) = −diag (0.004vx , 0.004vy , 0.01vz , 0.8ωθ , 0.8ωψ , 0.8ωϕ ) ;
⎛ ⎞
cosϕsinθcosψ + sinϕsinψ 0 0 0
⎜ sinϕsinθcosψ + cosϕsinψ 0 0 0 ⎟
⎜ ⎟
⎜
⎜ cosθcosψ 0 0 0 ⎟ ⎟;
B (X) = ⎜
⎜ 0 1 0 0 ⎟ ⎟
⎝ 0 0 1 0 ⎠
0 0 0 0.5
⎛ ⎞ ⎛ ⎞
U1 N1 + N 2 + N 3 + N 4
⎜ U2 ⎟ ⎜ −N1 − N2 + N3 + N4 ⎟
U (t) = ⎜ ⎟ ⎜
⎝ U3 ⎠ = ⎝ −N1 + N2 + N3 − N4 ⎠ ;
⎟

U4 N1 − N2 + N3 − N4
C = diag (0.08, 0.036, 1.14, 0.008, 0.008, 0.004) .

The functional task of visual monitoring of marine traffic determines the

high-precision stabilization of the quadrotor UAV in the horizontal plane and
the movement along a given trajectory followed by a “soft” landing. The
minor controlled coordinates of the yaw angles ϕ, roll ψ; and pitch θ are
controlled by three optimized PID controllers [38]. The trajectories concern-
ing the higher coordinates x, y, z are formed for the condition of the third zero
derivatives of the vector of coordinates. For the given stabilization trajectory
of the quadrotor UAV, the switching moments of the control functions T 1 , T 2 ,
T are determined.
The block diagram of the robust-optimal quadrotor UAV control system
(Figure 12.10) includes the following blocks: a) CAVCU is a block of
synthesizing the specified values of the orientation angles of a quadrotor
UAV; b) CSFU is a block of synthesizing the corrective signal of a robust
loop; c) SCU is a block of switching control actions; d) SW is a switching
key; e) Wcor is a transfer matrix function that can be obtained from (12.32)
considering linearization and stationarity of matrices A and B. The physical
and reference models are set with a mismatch condition within ± 15%. A
piecewise constant component (average hourly wind speed Vw = 10 m/s) and
a high-frequency component formed by “white noise” of a given intensity
cause the wind impact.
The results of the simulation of movement along a given quadrotor UAV
trajectory (Figures 12.11–12.13: error and velocity graphs for the higher
 12.5 Experiments 313

Figure 12.10 Simulink-based robust-optimal stabilization scheme for a quadrotor UAV on a

given trajectory

coordinates, as well as lower coordinates of angular orientation, respectively)

demonstrate sufficiently small values of control errors (less than 1%).
For comparative analysis, the control accuracy and energy consumption
for the control system under consideration with the optimization of the
trajectory of three higher coordinates of the motion and three PID controllers
(Figures 12.14, 12.15) are compared; the graphs of the higher coordinates
314 Algorithmic Procedures Synthesis of Robust-Optimal Control

Figure 12.11 Errors of the higher coordinates over time

Figure 12.12 Velocity of the higher coordinates over time

Figure 12.13 Lower coordinates of angular orientation over time

 12.5 Experiments 315

Figure 12.14 Comparison of control accuracy between optimal control with variable
structure and PID control over time

Figure 12.15 Comparison of energy consumption between optimal control with variable
structure and PID control over time

and control, respectively, are shown by means of the index xyz), and system
with optimization by two higher coordinates x, y and a block of four PID
controllers for the remaining four coordinates (the graphs shown by mean
of the index xy), are given. The simulation results demonstrate a decrease in
energy consumption (values are given in arbitrary units) of control for the
proposed control system.
The simulation results (Figures 12.16, 12.17, the red line inside the "tube"
shows the optimal trajectory) demonstrates the errors change in the form of
ńtubesż for different values of external uncontrolled wind disturbance.
316 Algorithmic Procedures Synthesis of Robust-Optimal Control

Figure 12.16 “Error tubes” of the main coordinates x, y, z; Vw = 10 m/s

Figure 12.17 “Error tubes” of the main coordinates x, y, z; Vw = 3 m/s

For example, the problem of high-precision control of a quadrotor UAV

under the conditions of wind disturbance arises during navigation and envi-
ronmental monitoring of the marine environment to expand the information
base for assessing the safety of navigation. The simulation results show
the robust stability of the control system of the quadrotor UAV to external
disturbances, the uncertainty of the mathematical model and the minimum
deviations of the current trajectory from the formed optimal trajectory of
 12.6 Conclusion 317

stabilization. For the circuitry implementation, a Simulink-oriented scheme

of a robust-optimal control system for a quadrotor UAV was formed.

12.6 Conclusion
Optimal control systems provide moving objects with the highest control
quality indicators in terms of speed, accuracy, and energy consumption. In the
case of multidimensional nonlinear systems, there are significant restrictions
on the applicability of classical optimal control methods associated with the
computational difficulties due to boundary value problems complex solutions,
the correctness of the formation of weight matrices of optimality criteria and
others.
To optimize nonlinear systems up to the sixth order, the proposed algorith-
mic procedure of the synthesis of variable structure system for different types
of moving objects based on preliminary construction of optimal trajectories of
stabilization of dynamic processes has been applied. The generated minimax
optimality conditions for various criteria allow to consider the constraints on
the control actions and solve the synthesis problem for the corresponding
functional control problems: maximum speed (for critical modes) and mini-
mum energy consumption (for technological modes). The uncertainty of the
external environment and the dynamic object requires robust correction of the
optimal system. The generated optimal trajectories and control determine the
formation of the additional robust control loop to compensate incomplete a
priori definiteness of the mathematical model and uncontrolled (unmeasured)
disturbances and noises.
The proposed procedure for the synthesis of robust-optimal systems has
been tested for several models of a nonlinear moving object and demonstrated
the versatility and engineering practicality, as well as efficiency under the
influence of external limited disturbances in terms of accuracy and con-
trol energy consumption. A number of algorithmic procedures for optimal
trajectories construction, switching moments determination, and synthesis
of control functions for multidimensional systems have been developed.
Controlling the mismatch between the trajectory of a physical object and
the calculated optimal trajectory of the reference model allows to form a
robust subsystem that provides invariance to incomplete information about
the dynamic system taking into account the values of the optimal control from
the reference system. The presented results of the simulation of quadrotor
UAV stabilization and maneuvering of a vessel under the conditions of
external disturbance and parametric noise demonstrate the provision of the
318 Algorithmic Procedures Synthesis of Robust-Optimal Control

necessary control accuracy in the vicinity of the formed optimal trajectory of

controlled coordinates.

References
[1] Krasovsky, N.N.: Some problems of the theory of stability of motion,
Phizmathlit, Moscow (1959) (in Russian)
[2] Feldbaum, A.A.: Theory Of Optimal Systems, Phizmathlit, Moscow
(1963) (in Russian)
[3] Kuntsevich, V.M.: Synthesis of Robust Optimal Adaptive Control Sys-
tems for Nonstationary Objects under Bounded Disturbances. In: Jour-
nal of Automation and Information Sciences, NY, Begell house inc., vol.
36, issue 3, pp.14–24 (2004)
[4] Kuntsevich, V.M., et al (Eds): Control Systems: Theory and Applica-
tions. Series in Automation, Control and Robotics, River Publishers,
Gistrup, Delft, (2018)
[5] Kondratenko, Y.P., Kuntsevich, V.M., Chikrii, A.A., Gubarev, V.F.
(Eds): Advanced Control Systems: Theory and Applications. Series in
Automation, Control and Robotics, River Publishers, Gistrup, (2021)
[6] Kondratenko, Y.P., Chikrii, A.A., Gubarev, V.F., Kacprzyk, J. (Eds):
Advanced Control Techniques in Complex Engineering Systems: The-
ory and Applications. Dedicated to Professor Vsevolod M. Kuntsevich.
Studies in Systems, Decision and Control, Vol. 203. Cham: Springer
Nature Switzerland AG (2019)
[7] Kuntsevich, V.M.: Estimation of Impact of Bounded Perturbations on
Nonlinear Descrete Systems. In: Kuntsevich, V.M. et al. (Eds). Control
Systems: Theory and Applications. Series in Automation, Control and
Robotics, River Publishers, Gistrup, Delft, pp. 3–15 (2018)
[8] Emelynov, S.V., Korovin, S.K.: New types of feedback, Nauka, Phiz-
mathlit, Moscow, (1997) (in Russian)
[9] Emelynov, S.V.: Automatic control systems of variable structure: syn-
thesis of scalar and vector systems by state and by output. In: Nonlinear
dynamics and control, Phizmathlit, Moscow, vol. 5, pp. 5–24 (2007) (in
Russian)
[10] Utkin, V.I.: Sliding Modes in Control and Optimization, Springer-
Verlag, (1992)
[11] Shtessel, Y.B., Moreno, J.A., Fridman, L.M.: Twisting sliding mode
control with adaptation: Lyapunov design, methodology and applica-
tion. In: Automatica, vol. 75, pp. 229–235 (2017)
 References 319

[12] Dorf, R.,. Bishop, R.: Modern Control Systems, LBZ, Moscow (2002)
(in Russian)
[13] Boychuk, L.M.: Method of structural synthesis of nonlinear automatic
control systems, Energia, Moscow (1971) (in Russian)
[14] Galiulin, A.S.: Methods for solving inverse problems of dynamics,
Moscow, Nauka, (1986) (in Russian)
[15] Krutko, P.D.: Robust Stable Structures of Controlled Systems of High
Dynamic Accuracy. Algorithms and dynamics of motion control of
model objects. In: Control theory and systems, no. 2, pp. 120–140
(2005) (in Russian)
[16] Krutko, P.D.: Inverse problems of dynamics in the theory of automatic
control, Mashinostroenie, Moscow, (2004) (in Russian)
[17] Kondratenko, Y.P., Timchenko,V.L.: Optimal feedback switching
method for linear control system. In: Intern. System and Networks:
Mathematical theory and applications, Berlin, Academie Verlag, Vol. 2,
pp. 291–292 (1994)
[18] Timchenko, V.L., Lebedev, D.O., Robust-optimal stabilization of non-
linear dynamic systems. In: Radio Electronics, Computer Science,
Control, no. 3., pp. 185–197 (2019)
[19] Gabasov, R., Kirillova, F.M., Ruzhitskay, E.A.: Implementation of
Bounded Feedback in a Nonlinear ontrol Problem. In: Cybernetics and
Systems Analysis, no. 1, pp. 108–116 (2009) (in Russian)
[20] Balashevich, N.V., Gabasov, R., Kalinin, A.I. and Kirillova, F.M.: Opti-
mal Control of Nonlinear Systems. In: Computational Mathematics and
Mathematical Physics, vol. 42, no. 7, pp. 931–956 (2002)
[21] Horowitz, I.M.: Survey of quantitative feedback theory (QFT). In: Int.
Journal of Robust and Non-Linear Control, vol. 11., no. 10., pp. 887–921
(2001)
[22] Larin, V.B.: On the reversal of the problem of analytical design of
controllers. In: Control and Information Promblems, no. 1, pp. 17–25
(2004) (in Russian)
[23] Yaesh, I., Boyarski, S., Shaked U.: Probability-guaranteed robust H∞
performance analysis and state-feedback design. In: Systems & Control
Letters, vol, 48, issue 5, pp. 351–364 (2003)
[24] Chesnov, V.N.: Synthesis of H∞ - controllers for Multidimensional
Systems of Given Accuracy and Stability. In: Autom. Remote Control,
vol. 72, no. 10, pp. 2161–2175 (2011)
[25] Roberts, G.N.: Trends Marine Control Systems. Im: IFAC, CAMS’07,
Zagreb, Croatia (2007)
320 Algorithmic Procedures Synthesis of Robust-Optimal Control

[26] Ornerdi, c E., Robert,s G.N.: A fuzzy track-keeping autopilot for ship
steering. In: Journal of Marine Engineering and technology, London,
no. A2, pp. 23–35 (2003)
[27] Lukomsky, Y.K., Chugunov, V.S.: Marine mobile objects control sys-
tems, L., Sudostroenie, (1988) (in Russian)
[28] Fossen, T.I., Perez, T.: Kinematic models for maneuvering and seakeep-
ing of marine vessels. In. Journal of modeling, identification and control,
vol. 28, issue 1, pp. 19–30 (2007)
[29] Johansen, T., Fossen, T.I.: Control allocation—A survey. In. Automat-
ica, vol.49, pp. 1087–1103 (2013)
[30] Comasòlivas, R., Escobet, T. and Quevedo, J.: Automatic design of
robust PID controllers based on QFT specifications. In: Proceeding of
IFAC Conference on Advances in PID Control, pp. 715–720 (2012)
[31] Timchenko, V.L., Kondratenko, Y.P.: Robust stabilization of marine
mobile objects on the basis of systems with variable structure of feed-
backs. In: Intern. Journal of Automation and Information Sciences, NY,
Begell house inc., vol. 43, issue 6, pp. 16–29 (2011)
[32] Timchenko, V.L.: Synthesis of variable structure systems for stabi-
lization of ships at incomplete controllability. In: Intern. Journal of
Automation and Information Sciences, NY, Begell house inc., vol. 44,
issue 6, pp. 8–19 (2012)
[33] Timchenko, V.L., Ukhin,O.A.: Optimization of Stabilization Processes
of Marine Mobile Object in Dynamic Positioning Mode. In: Intern.
Journal of Automation and Information Sciences, NY, Begell house inc.,
vol. 46, issue 7, pp. 40–52 (2014)
[34] Timchenko, V.L., Ukhin, O.A.: Variable Structure Robust-Optimal Sys-
tems for Control of Marine Vehicles. In: Proceeding of the IEEE 4th
International Conference MSNMC, Kyiv, Ukraine, pp. 151–155 (2016)
[35] Timchenko, V.L., Lebedev, D.O.: Automated Algorithmic Procedure of
Robust-Optimal Control System’s Synthesis for Marine Vehicles. In:
Proceeding of the IEEE 4th International Conference MSNMC, Kyiv,
Ukraine„ pp. 98–101 (2018)
[36] Timchenko, V.L., Ukhin, O.A. and Lebedev, D.O.: Optimization of
nonlinear systems of variable structure for control of marine moving
vehicles. In: Intern. Journal of Automation and Information Sciences,
NY, Begell house inc., vol. 49, issue 7, pp. 33–47 (2017)
[37] Altug, E., Ostrowski, J.P. and Taylor. C.J.: Control of a quadrotor heli-
copter using dual cameravisual feedback. In: The International Journal
of Robotics Research, vol. 24, no. 5, pp. 329–341 (2005)
 References 321

[38] Pyrkin, A.A., Maltseva, T.A., Labadin, D.V., Surov, M.O. and Bobtsov,
A.A.: Sintez sistemy upravleniya kvadrokopterom s ispol’zovaniyem
uproshchennoy matematicheskoy modeli. In: Izv. Vuzov. Priborostroye-
niye, RF, SPb, vol. 56, no. 4, pp. 47–51 (2012) (in Russian).
[39] Andreev, M.A., Miller, B., Miller, B.M. and Stepanyan, K.V.: Path
planning for unmanned aerial vehicle under complicated conditions and
hazards. In: International Journal of Computer and Systems Sciences,
vol. 51, no. 2, pp. 328–338 (2012)
[40] Younes, Y.A., Drak, A., Noura, H. et al.: Robust Model-Free Control
Applied to a Quadrotor. In: J. Intell. Robotics Syst., 84, pp. 37–52 (2016)
[41] Larin, V.B., Tunik, A.A.: Synthesis of the quad-rotor flight control sys-
tem. In: Proceeding of the IEEE 4th International Conference MSNMC,
Kyiv, Ukraine, pp. 12–17 (2016)
[42] Wang, D., Liu, D., Li, H., et al.: An Approximate Optimal Control
Approach for Robust Stabilization of a Class of Discrete-Time Non-
linear Systems With Uncertainties. In: IEEE Transactions on Systems,
Man, and Cybernetics: Systems, vol. 46, no. 5, pp. 713–717 (2016)
[43] Lu, T., Wen, P.: Time Optimal and Robust Control of Twin Rotor
System. In: Proceeding of the IEEE International Conference on Control
and Automation, Guangzhou, China, pp. 862–866 (2007)
[44] Faruque, I.A., Muijres, F.T., Macfarlane, K.M. et al.: Identification of
optimal feedback control rules from micro-quadrotor and insect flight
trajectories. In: Biol Cybern., 112, pp. 165–179 (2018)
[45] Kamran, J.M., Ahmed, S.F., Bakar, M.I., Ali, A.: Horizontal Motion
Control of Underactuated Quadrotor Under Disturbed and Noisy
Circumoyostances, Information and Communication Technology. In:
Advances in Intelligent Systems and Computing, Springer, vol. 625.,
pp. 63–79 (2018)
[46] Jayakrishnan, H.J.Z: Position and Attitude control of a Quadrotor UAV
using Super Twisting Sliding Mode. In: IFAC-PapersOnLine, vol. 49,
issue 1, pp. 284–289 (2016)
[47] Moreno, J.A., Osorio, M.: A Lyapunov approach to second-order sliding
mode controller and observers. In: Proceeding of the 47th Conference on
Decision and Control, pp. 2856–2861 (2008)
[48] Matveev, A.S., Teimoori, H., Savkin, A.V.: A method for guidance
and control of an autonomous vehicle in problems of border patrolling
and obstacle avoidance. In: Automatica, vol. 47, issue 3, pp. 515–524
(2011)
322 Algorithmic Procedures Synthesis of Robust-Optimal Control

[49] Leira, F.S., Johansen, T.A. and Fossen,T.I.: Automatic detection, classi-
fication and tracking of objects in the ocean surface from UAVs using a
thermal camera. In: Proceeding of Aerospace Conf., pp. 1–10(2015)
[50] Liu, Y., Chen, C., Wu, H. et al.: Structural stability analysis and opti-
mization of the quadrotor unmanned aerial vehicles via the concept of
Lyapunov exponents. In: Int J Adv Manuf Technol 94, pp. 3217–3227
(2018)
[51] Liu, Y., Li X., Wang, T., Zhang, Y., Mei, P.: The Stability Analysis
of Quadrotor Unmanned Aerial Vechicles. In: Wearable Sensors and
Robots. Lecture Notes in Electrical Engineering, Springer, vol. 399,
pp. 383–394 (2017)
[52] Liu, Y., Li, X., Wang, T., et al.: Quantitative stability of quadrotor
unmanned aerial vehicles. In: Nonlinear Dyn, pp. 1819–1833 (2017)
[53] Gharib, M. R., and Moavenian, M.: Full dynamics and control of a
quadrotor using quantitative feedback theory. In:Int. J. Numer. Model,
29, pp. 501–519 (2017)
[54] Xu, Y., Tong, Ch.: Quantitative feedback control of a quadrotor. In:
Proceedings of the IEEE International Symposium on Industrial Elec-
tronics, Hangzhou, China, pp. 1309–1314 (2012)
[55] Raffo, G. V., Ortega M. G., Rubio F. R.: An integral predictive/nonlinear
H∞ control structure for a quadrotor helicopter. In: Automatica, vol. 46,
issue 1, pp. 29–39 (2010)
[56] Viegas, D., Batista, P., Oliveir,a.P., Silvestre, C.: Decentralized H2
observers for position and velocity estimation in vehicle formations
with fixed topologies. In: Systems & Control Letters, vol. 61, issue 3,
pp. 443–453 (2012)
[57] Falkenberg, O., Witt, J., Pilz, U., Weltin, U., Werner, H.: Model Identi-
cation and H∞ Attitude Control for Quadrotor MAV’s. In: ICIRA, Part
II, pp. 460–471 (2012)
[58] Sushchenko, O.A., Shyrokyi, O.V.: H 2 /H-Optimization of system for
stabilization and control by line-of-sight orientation of devices operated
at UAV. In: Proceedings of the IEEE 3rd International Conference
APUAVD, pp. 235–238 (2015)
[59] Zhang, X., Kamgarpou,r M., et al.: Robust optimal control with
adjustable uncertainty sets. In: Automatica, vol. 75, pp. 249–259 (2017)
[60] Sushchenko, O.A.: Robust control of angular motion of platform with
payload based on H∞ -synthesis. In: Journal of Automation and Infor-
mation Sciences, vol. 48, issue 12, pp. 13–26 (2016)
 References 323

[61] Satici, A.C., Poonawala, H., Spong, M.W.: Robust Optimal Control of
Quadrotor UAVs. In: IEEE Access, vol. 1, pp. 79–93 (2013)
[62] Vossen, G., Maurer, H.: On L1-minimization in optimal control
and applications to robotics. In: Optim. Control Appl. Meth., 27,
pp. 301–321 (2006)
[63] Chen, Y., Han, J., & Zhao, X., Three-dimensional path planning for
unmanned aerial vehicle based on linear programming. In: Robotica,
30(5), pp. 773–781 (2012).
[64] Jiang, B., Bishop, A.N., Anderson, B. D.., Drake S.P.: Optimal path
planning and sensor placement for mobile target detection. In: Auto-
matica, Vol. 60, pp. 127–139 (2015)
[65] Lewis, F.L., Vrabie, D., Vamvoudakis, K. G.: Reinforcement Learning
and Feedback Control: Using Natural Decision Methods to Design.
Optimal Adaptive Controllers. In: IEEE Control Systems Magazine,
vol. 32, no. 6, pp. 76–105 (2012)
[66] Bisheban, M., Lee T.: Geometric Adaptive Control With Neural Net-
works for a Quadrotor in Wind Fields. In: Proceeding of the IEEE
Transactions on Control Systems Technology, pp. 1–19 (2019)
[67] Timchenko, V.L., Lebedev, D.O., Kuklina, K.A. and Timchenko, I.V.:
Robust-optimal control system of quad for maritime traffic’s monitoring.
In: Proceeding of the IEEE 4th international conference APUAVD,
Kyiv, Ukraine„ pp. 192–196 (2017)
[68] Timchenko, V.L., Lebedev, D.O., Optimization of Processes of Robust
Control of Quadcopter for Monitoring of Sea Waters. In: Journal of
Automation and Information Sciences, NY., Begell house, vol. 51,
issue. 2, pp. 1–10 (2019)
[69] Timchenko, V.L., Lebedev, D.O.: Robust control systems with subopti-
mal models for stabilization of UAV. In: Proceedings of the 2019 IEEE
5th International Conference APUAVD, Kyiv, Ukraine, pp. 124–129
(2019)
[70] Timchenko, V.L., Lebedev, D.O.: Synthesis of Optimal Phase Trajec-
tory for Quadrotor UAV Stabilization. In: Proceedings of the IEEE 6th
International Conference MSNMC, Kyiv, Ukraine, pp. 20–24 (2020)
[71] Sumacheva, V.A., Kharitonov, V.L.: Calculation of the H2 Norm of the
Transfer Matrix of a System of Neutral Type. In: Differential Equations
And Control Processes, vol. 4, pp. 22–32 (2014)
 Part III
Recent Developments in Collaborative Automation
 13
Modeling of Cyber-Physical Systems

Illya Holovatenko, and Andrii Pysarenko

Department of Information Systems and Technologies, National Technical

University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”,
Kyiv, Ukraine
E-mail: [email protected]; [email protected]

Abstract
The purpose is to review current research in the field of constructing cyber–
physical systems. The structure of the chapter is as follows: a review of
the most effective approaches and methods of constructing cyber–physical
systems; using these methods—design cyber–physical–logistical system; per-
form the review of the obtained results and come up with a plan for further
researches based on the reviewed methods. For validation of the designed
system we came up with a live scenario from the logistical systems: the
autonomous object delivers goods from one warehouse to another. We also
will model obstacles along the way. The system should proactively react to
this and recalculate the optimal path. This will lead to the timely delivery of
goods to the destination.

Keywords: Cyber–physical system, logistics, optimal control, route

planning.

13.1 Introduction
Cyber–physical systems are the integration of computational and physical
processes. Controllers and computer networks control physical processes,
usually through feedback, which is also affected by physical processes and
vice versa.

327
328 Modeling of Cyber-Physical Systems

The economic and social potential of such systems is much greater than
that of their counterparts, which is why the main global investment goes to the
development of this technology. Nevertheless, there are significant problems,
especially because the physical components of such systems require certain
security and reliability requirements that are qualitatively different from those
for general purpose computing.
Logistics is a set of entities involved in the delivery of goods and
services to the end customer. Logistics management is the effective end-to-
end process management, which includes: 1) planning the optimal delivery
route; 2) optimal appointment of “performers” (autonomous objects that carry
out delivery); 3) coordination of system subjects among themselves; 4) risk
forecasting; 5) timely response to emergency situations. The logistics system
faces problems that can potentially lead to inefficiencies and waste in supply
chains, such as delayed deliveries, rising fuel costs, inconsistent suppliers,
and ever-increasing end-user expectations.

13.2 Review of Modeling Methods for Cyber–Physical

Systems
One of the approaches to modeling cyber–physical systems is their repre-
sentation in the form of templates based on FCA (formal concept analysis)
to optimize the interoperability of cyber–physical systems in Industry 4.0
[1]. Interoperability can be defined as the ability of two or more systems to
exchange, understand and consume information [10].
A cyber–physical system is presented in the form of a certain metamodel
of the interaction of the nth number of cyber–physical systems. The compo-
nents of a cyber–physical system are sets of physical and cyber components.
A cyber–physical system is a structural agglomeration of these elements,
which can also include other cyber–physical subsystems, forming a new
cyber–physical system.
There are two relationships of different nature between the components
of a cyber–physical system:
1) The relationship between the physical components of cyber–physical
subsystems, implying the transfer of any kind of physical object.
2) The relationship of communication between cyber components, which
means the presence of information and/or control channel between the
components.
The components of the system perform certain functions depending on their
role in the system, and following this, they have some input and output signals
 13.2 Review of Modeling Methods for Cyber–Physical Systems 329

that capture information flows between this element and the elements with
which it is associated with the help of the above relations.
The definition of the system, in this case, will be as follows. Let there be
a system
{P, C, CPS, Rp , Rc } (13.1)
where P is the set of physical components that are related by the relation-
ship Rp ;
1) C—cyber components that are related by the relationship Rc ;
2) CPS—a cyber–physical system, including as a child, having the same
properties as the parent.
Figure 13.2 shows an example of building a composite cyber–physical sys-
tem. It includes three child cyber–physical systems (CPS1, CPS2, CPS3) as
part of one parent (CPS4); each of the systems has at least one physical (P)
and one cyber component (C); communication occurs only between com-
ponents of the same type (physical component physical component, cyber
component ⇔ cyber component).

Figure 13.2 Example of a composite cyber–physical system

330 Modeling of Cyber-Physical Systems

The resulting metamodel of a cyber–physical system is further described

in UML notation as shown in Figure 13.3.

Figure 13.3 Metamodel in UML notation

 13.2 Review of Modeling Methods for Cyber–Physical Systems 331

The main focus is on the “form a single CPS” relationship between the
“Cyber Component” and “Physical Component” classes, which relate to each
other as “one to one” (one cyber component to one physical component).
As stated in the metamodel, the “Cyber Component” and “Physical Com-
ponent” classes are subclasses of the “CPS Component” base class, which
contains the fundamental properties. Each “CPS Component” can either act
as a communicator between two components, or be an independent unit and
communicate with other homogeneous components. Communication is done
via input and output interfaces.
The next approach to the description of cyber–physical systems is the use
of Business Process Modeling Notation (BPMN 2.0), taking into account the
specifics of cyber–physical systems. Classifying new metaclasses requires a
deep understanding of BPMN metaclasses [2, 11–14].
For this, new abstract classes of cyber–physical systems are introduced as
a set of BPMN subclasses. This concept will allow the developer of a cyber–
physical system to describe the logic of internal processes, various types of
actions inside and real objects.
In a cyber–physical system, business processes must be defined accord-
ing to a centralized approach to execute business logic. BPMN defines the
behavior of different participants using different pools.
The first pool is for physical processes. The second pool is for cyber pro-
cesses. The third pool is a central controller that directly coordinates the inter-
action between cyber and physical processes. Therefore, the BPMN4CPS
process must consist of at least three pools: physical, controller, and cyber.
The separation into the processing logic, executed at the physical level,
the controller and the cyber part, forces the modeler to explicitly indicate the
interaction between them using messages. Data objects provide information
about what physical actions need to be performed and/or what information
they generate. Data objects are the temporarily stored data of a running
process instance.
Figure 13.4 shows an example illustrating the execution of a closed loop
of a cyber–physical task in the context of three pools: physical, controller,
and cyber process pools.
In cyber–physical systems, there are three types of tasks: cyber, man-
ual, and physical. These types of actions are also covered in the BPMN
metamodel.
Cybertask is an extension of the standard BPMN service task. This task is
performed at the cyber level and, in essence, is the support for the controller
in the formation of the control action. For example, solving optimization
332 Modeling of Cyber-Physical Systems

Figure 13.4 A closed-loop of a cyber–physical system

problems, the cyber level finds a solution to a certain problem, which it sends
to the controller.
A physical task is an atomic activity within a process. The physical
process can be divided into two parts:
1) retrieve sensor data and send them to the controller level;
2) receipt by executive devices of control actions from the controller.
In this model, the controller plays the role of a kind of “mediator” and
“aggregator”. The role of the “mediator” is that the controller receives data
from the sensors and the cyber layer and sends it to the physical layer.
 13.2 Review of Modeling Methods for Cyber–Physical Systems 333

The role of the “aggregator” is that having stabilized the system in the event
of a disturbance and having received data from the cyber level, the controller
brings this data into a single form that is understandable for executive devices.
The third approach to modeling cyber–physical systems is the model-
based design method for constructing dynamic systems [3, 15–17].
Designing a complex cyber–physical system, especially with heteroge-
neous subsystems distributed over networks, is challenging. Commonly used
design methods are complex and include mathematical modeling of physical
systems, formal computation models, modeling of heterogeneous systems,
software synthesis, verification, validation, and testing.
The proposed algorithm consists of 10 steps:
1) problem statement;
2) description of physical processes by equations;
3) description of minimum/maximum parameters and/or restrictions;
4) synthesis of control;
5) selection of computing software;
6) choice of hardware;
7) modeling of the process;
8) design;
9) software development;
10) performance check and testing.
Let’s briefly describe each step of the algorithm.
Problem statement.
A preparatory step that requires maximum attention. In this step, a basic
understanding of the problem is formed. For the most part, this step involves a
written description of the problem, without the use of complex mathematics.
Description of physical processes by equations.
In this step, an understanding of the relevant physical systems is formed,
such as the environment in which the cyber–physical system is located, or
the physical processes that need to be controlled.
Models of physical processes are simplified representations of real
systems and usually take the form of systems of differential equations.
Description of minimum/maximum parameters and/or restrictions;
At this stage, need to highlight the constants, adjustable parameters, and
variables that need to be controlled.
334 Modeling of Cyber-Physical Systems

Also need to determine the values that characterize the system and/or
physical processes, such as minimum and maximum values, initial states,
final states.
Synthesis of control.
This step requires determining the conditions under which the physical pro-
cesses are controllable and synthesizing the control law. It is also necessary
to define the parameters of time delays, sampling rate, and quantization
so that the physical dynamics of interest can be accurately measured and
appropriately controlled.
Selection of computing software.
The formal computation model defines semantics that often leads to greater
analyzability and the ability to model cyber–physical systems using hetero-
geneous modeling tools.
Models described by formal computation models are easier to analyze in
terms of determinism, runtime, state reachability, and memory usage.
Choice of hardware.
When having a basic understanding of all the components of the system, it is
necessary to select the components from which the production model will be
built.
The selection of equipment should be based on the fact that the finished
model should adequately interact with the simulated physical systems and
implement the control law in a high-precision measure.
For each component, consider its input and output bandwidth, input-to-
output latency, power consumption, resolution, and measurement speed, as
well as mechanical parameters such as form factor, electrical noise rejection,
durability, and lifespan.
Modeling of the process.
This step allows us to verify the adequacy of the developed model of the
cyber–physical system. It is necessary to collect quantitative and qualitative
parameters to make sure that the optimal control law takes the system from
the initial state to the final one. Should also check that all limits and ranges
for each module are met.
Design.
This involves assembling the production model itself according to the spec-
ifications. It is best to build a model so that tests can be run at each step
 13.3 Problem Formulation 335

to ensure that the individual components and subsystems are performing as

planned. This, in turn, facilitates the transition from simulation to testing.
Software development.
The creation of the controller software is a critical step in this process.
The resulting model of the system in the previous steps, in some modeling
environments, can be easily converted into executable code.
If the synthesis of the code is not possible the handwritten code must
carefully follow the chosen computation models. Code execution must be
verified on a real object because code generators and compilers can introduce
software artifacts.
Performance check and testing.
Computing systems can be isolated from physical systems through hardware
testing, while programmable hardware, such as controllers or FPGAs, simu-
lates feedback from physical or other computational processes. Runtime and
latency measurements can be used to refine previous models, and test results
can indicate errors in modeling or implementation.
Formal validation and validation provide insight into the behavior of an
algorithm over some or all combinations of its inputs or over time.

13.3 Problem Formulation

The concept of logistics includes several components: goods; objects, that
deliver goods (trucks, ships, planes, trains); and places where goods are
delivered (end consumer, intermediate warehouse, loading/unloading point,
ports, railway junctions).
Let’s designate each of the elements. Objects that deliver goods (ODG),
places where goods are delivered (PGD). Now the logistics management
systems receive information on the status of the cargo from PGD. These
systems do not reflect what happens between individual PGDs. In the event
of an emergency on the ODG route (the ship got into a storm, the truck got
stuck in a traffic jam), the system will not display this information and will
not generate the appropriate control action to optimize the delivery of the
cargo.
Let us introduce a system of connections within this system:
1) communication “ODG” – “PGD”;
2) communication “ODG” – “ODG”;
3) communication “PGD” – “PGD”;
336 Modeling of Cyber-Physical Systems

Figure 13.5 An example of inter-component interaction within a cyber–physical system

4) Communication “System” – “PGD”;

5) Communication “System” – “ODG”.
Consider the following cyber–physical system:
1) ODG received a logistic route for the delivery of cargo from PGD1 to
PGD2 and an algorithm for optimal, in terms of energy consumption,
control from a centralized cloud server;
2) ODG begins to move, in parallel tracking external disturbance on the
route (for example, using a radar);
3) In the event of a situation in which move along the planned route
is impossible, ODG sends the relevant information to the centralized
cloud;
4) The cloud, in turn, recalculates the logistic route and the algorithm for
the optimal control in terms of energy consumption;
5) Having received the updated data, ODG continues its journey to PGD2.
Such a dynamic response to changes in the external environment will min-
imize the problem of untimely delivery of cargo. Figure 13.5 shows an
example of how ODG communicates with the cloud.

13.4 Building Models of a Logistic Cyber–Physical

System
We use the cyber–physical system metamodel method. Figure 13.6 shows
the inter-component interaction of a cyber–physical system, proposed in the
formulation of the problem.
An instance of a cyber–physical system class is denoted as a compound
ellipse, where the solid half indicates the physical component and the dotted
 13.4 Building Models of a Logistic Cyber–Physical System 337

Figure 13.6 The metamodel of a cyber–physical system

half indicates the cyber component. Accordingly, the “form a single CPS”
relation is represented by combining the half ellipses into a whole ellipse.
Some properties of a cyber–physical system can be directly deduced from
Figure 13.6 by the type of connection between cyber–physical components.
For example, a cyber–physical component that has no physical inputs and
outputs and only interacts in the cybernetic dimension can be thought of as a
computing node. In this notation, “System” is a computational node.
In addition, a component that receives only physical input and provides
only cyber output can be called a sensor. A radar that interacts with ODG
through physical signals and provides a cyber response on the state of the
road. In the metamodel, it can be referred to as sensors.
Figure 13.7 shows the implementation of a model of a logistics cyber–
physical system in BPMN4CPS terminology.
Here, the extended BPMN notation divides the processing logic into three
parts: the cyber part, the controller, and the physical part.
These parts interact with each other through the exchange of messages.
The set of components here is as follows
1) ODG;
2) radar installed on the ODG;
338 Modeling of Cyber-Physical Systems

Figure 13.7 BPMN4CPS cyber–physical System Notation

3) controller;
4) the cyber component of the system responsible for data processing
(cloud computing resources).
This notation clearly shows the relationship between components. ODG is a
basic component.
The level of the physical process is a radar as a sensor device and a
steering wheel system as an executive component. The radar tracks objects
in the environment and the distance to them. Having received information
about the next object from the radar, the controller sends data to the cloud
 13.4 Building Models of a Logistic Cyber–Physical System 339

computing center. The result of data processing by the cloud computing

center is sent to the controller, which transmits control actions to the input
of the steering wheel control system to change the direction and/or speed of
movement, according to the current situation.
The level of controller interaction is a set of hardware. The main task
of this layer is to reconcile data between the physical and cyber component
layers. Having received data from the radar, the controller must bring them
into an acceptable format for a computing resource. The controller performs
a similar function, having received data from a computing resource, with only
one difference, which brings the data into a format that is acceptable for an
executive device at the physical level.
And finally, the level of cyber processes. The main component of this
layer is the cloud-computing resource. Receiving data for analysis, namely
the state of the environment received from the radar, a decision is made
whether it is necessary to recalculate the logistic route and/or the optimal
control coefficients for the further following to the PGD. If necessary, the
corresponding algorithms perform the calculation of the refined route and the
synthesis of the optimal traffic control. Subsequently, this data is sent to the
controller level for subsequent transmission to the actuators.
Now we turn to the construction of the model by the method pro-
posed in [3].
Problem statement.
Consider the following problem:
1) ODG received a logistic route for cargo delivery from PGD1 to PGD2
and an algorithm for optimal, in terms of energy consumption, control
from a centralized cloud server;
2) ODG begins to move, in parallel tracking external influences on the route
(for example, using a radar);
3) In the event of a situation in which move along the planned route
is impossible, ODG sends the relevant information to the centralized
cloud;
4) The cloud, in turn, recalculates the logistic route and the algorithm for
the optimal control in terms of energy consumption;
5) Having received the updated data, ODG continues its journey to PGD2 .

Description of physical processes by equations.

Table 13.1 describes the main physical definitions and constants of ODG.
340 Modeling of Cyber-Physical Systems

Table 13.1 Physical definitions of ODG

Symbol Notation
L ODG width
rr Right wheel radius
rl Left wheel radius
ωr Right wheel angular velocity
ωl Left wheel angular velocity
Vr Right wheel velocity
Vl Left wheel velocity
ω Overall angular velocity
θ The rotation angle of the ODG
ϕ The rotation angle of the wheel

Let ϕ be the angle of rotation of the wheel. It is known from kinematics

that the angular velocity is expressed by the following equation:
dϕ
ω= (13.9)
dt
For the right wheel, the angular velocity is:
dϕr
ωr = . (13.10)
dt
The angular velocity of the left wheel is:
dϕl
ωl = . (13.11)
dt
Having both components of the angular velocity of the wheels, it is
possible to express the total angular velocity of the ODG, ωODG :
ωr − ωl
ωODG = r . (13.12)
L
From the angular velocity it is possible to determine the overall linear
velocity of the ODG, VODG :
ωr + ωl
VODG = r . (13.13)
2
Then the projections of the velocity can be defined as:

Vx = V cos(ω) = x (13.14)

Vy = V sin(ω) = y (13.15)
 13.4 Building Models of a Logistic Cyber–Physical System 341

Description of minimum/maximum parameters and/or restrictions.

The main problem of logistics is the late delivery of goods. Based on this,
the first limitation is introduced: the final time of cargo delivery from PGD1
to PGD2 . It should also be mentioned that it is not possible to achieve full
compliance with the time frame. The obvious proof of this is that proactively
responding to environmental changes and recalculating the logistics route
also takes a certain amount of time. So, it is rational to introduce admissible
boundary values for the final delivery time of the cargo.
The movement of the ODG along the route is the main task, but one
should also keep in mind that significant financial costs of logistics are the
cost of fuel or electricity consumed by the ODG. This leads to the second
criterion—the optimality of following the ODG in terms of fuel or electricity
consumption.
Synthesis of control.
Control synthesis is carried out by the approach described in [4, 18].
Table 13.2 lists the conventions and their descriptions that will be used below.
Consider a linear model [5, 6]:


x(t) = A(t)x(t) + B(t)u(t), (13.17)
y(t) = C(t)x(t).

Table 13.2 Optimal tracking system notation

Symbol Notation
A,B,C,D State-space matrices
e Difference between desirable and real outputs
F The terminal constant of criterion
H Hamiltonian
J Optimal criterion
K Solution of the Riccati equation
p Additional vector function
Q,R Weighting matrices of criterion
u Control law
x Vector of states of the state-space model
y Vector of outputs of the state-space model
z Desired output
342 Modeling of Cyber-Physical Systems

And the energy-optimal consumption criterion:


1 T 1 T T
J = e (T )Fe(T ) + [e (t)Q(t)e(t) + uT (t)R(t)u(t)]dt (13.18)
2 2 t0

where
e(t) = z(t) − y(t) (13.19)
e(t) – difference between desirable z(t) and real y(t) outputs. Rewrite e(t)
as a function of z(t) and x(t):

e(t) = z(t) − C(t)x(t) (13.20)

Substitute the error (13.19) into the optimality criterion (13.18) and find
the Hamiltonian H of the system:
1 1
H = [z(t) − C(t)x(t)]T Q(t)[z(t) − C(t)x(t)] + uT (t)R(t)u(t)p0 (t)
2 2
+ A(t)T x(t)T p(t) + B(t)T u(t)T p(t). (13.21)

The additional vector function p(t) is described by:

⎡ ⎤
p0 (t)
⎢ p1 (t) ⎥
p(t) = ⎢⎣
⎥
⎦ (13.22)
...
pn+1 (t)
so, 8 9T
p(T ) = −1 0 ... 0 (13.23)
and can be found from the following equation:
 ∂H
p(t) = − (13.26)
∂x(t)
Since there are no control restrictions:
∂H
max H : =0 (13.27)
u ∂u(t)
so,
∂H
= R(t)u(t) + BT (t)p(t) = 0 (13.28)
∂u(t)
 13.4 Building Models of a Logistic Cyber–Physical System 343

where optimal control is written as:

u(t) = −R−1 (t)BT (t)p(t) (13.29)
Differentiate (13.26) and obtain:

p(t) = −CT (t)Q(t)C(t)x(t) − AT (t)p(t) + CT (t)Q(t)z(t) (13.30)
Substitute the resulting control (13.29) into the state Equation (13.17):

x(t) = A(t)x(t) − B(t)R−1 (t)BT (t)p(t) (13.31)
Let the relationship between x(t) and p(t) be written by the following
equation:
p(t) = K(t)x(t) − g(t). (13.32)
Therefore, state Equation (13.31) can be rewritten as:

x(t) = [A(t) − S(t)K(t)]x(t) + S(t)g(t) (13.33)
where,
S(t) = B(t)R−1 (t)BT (t) (13.34)
Differentiate the vector function p(t) in (13.32):
   
p(t) = K(t)x(t) + K(t) x(t) − g(t) (13.35)
and substitute (13.33) in the (13.35):
  
p(t) = [K(t) + K(t)A(t) − K(t)S(t)K(t)]x(t) + K(t)S(t)g(t) − g(t)
(13.36)
Substitute (13.32) into the equation of optimal control (13.29) and obtain:
u(t) = −R−1 (t)BT (t)[K(t)x(t) − g(t)] (13.37)
Where K(t) – is a real, symmetric, positively defined matrix of dimension
n × n, which is the solution of the obtained Riccati equation:
K(t) = −K(t)A(t) − AT (t)K(t) + K(t)B(t)B−1 (t)BT (t)K(t)
− CT (t)Q(t)C(t) (13.38)
with a boundary condition:
K(T ) = CT (T )FC(T ) (13.39)
344 Modeling of Cyber-Physical Systems

The vector g(t) (with n components) is the result of the solution of the
differential equation:

g(t) = −[A(t) − B(t)R−1 (t)BT (t)K(t)]T g(t) − CT (t)Q(t)z(t) (13.40)

with an appropriate boundary condition:

g(T ) = CT (T )Fz(T ) (13.41)
Therefore, by solving the Riccati Equation (13.38) find the K(t) matrix also,
solving the differential Equation (13.40) in the reverse time will give the value
of the entire vector g(t). Substituting the obtained values of K(t) and g(t)
into Equation (13.37), obtain the optimal input control of the system, which
minimizes criterion (13.18) and solves the tracking problem.
Selection of computing software.
Modeling of the physical component of the system will be performed in the
MATLAB package using the Automated Driving Toolbox [7] add-on. The
Automated Driving Toolbox provides algorithms and tools for the design,
simulation, and testing of advanced driver-assistance systems (ADAS) and
autonomous driving. It is also possible to design and test computer vision
systems, lidar/radar sensing systems, in addition to path planning capabilities
and the development of vehicle controllers.
Choice of hardware.
Based on the task at hand, a cyber–physical system must have a certain
computing center, which can be represented as a cloud server. The cloud
provider can be Microsoft Azure. For the functioning of a cyber-component
system, it is enough to have a server with the following parameters:
1) Linux operating system;
2) 8 GB of RAM;
3) 4 virtual processor cores;
4) 64 GB of storage.
These parameters correspond to the virtual machine of the F4s v2 series [8]
optimized for computing.
Modeling of the process.
MATLAB is the main tool for modeling a scenario that suits the task at hand.
First, using the Automated Driving Toolbox is easier with MATLAB. Sec-
ondly, modeling of cyber–physical interaction through a computing “cloud”
 13.5 Simulation Results 345

can be achieved through MATLAB Instrument Control Toolbox [9]. Based

on this, using the MATLAB ecosystem for modeling is preferable.
The subsequent steps “Design”, “Software development” and “Perfor-
mance check and testing” are associated with the real implementation of the
system and require a real self-driving ODG traveling from PGD1 to PGD2 .
For research purposes, these steps will be skipped and the results will be
collected from the “Modeling of the process” step.

13.5 Simulation Results

Figure 13.8 shows the initial logistic route from PGD1 to PGD2 as received
by the ODG from the centralized cloud.

Figure 13.8 Primary trajectory obtained by ODG at the initial moment of time
346 Modeling of Cyber-Physical Systems

Figure 13.9 Script for constructing the primary trajectory

The algorithm for constructing the primary trajectory written in MAT-

LAB is shown in Figure 13.9.
The script consists of three parts:
1) building a road model using the Automated Driving Toolbox;
2) converting the road model into the so-called cost map;
3) the last step is to find the trajectory between two points using the RRT
(Rapidly-exploring random tree) algorithm.
Then the ODG began following the logistic route. While driving, external
influences on the roads, such as traffic jams or obstacles, are continuously
monitored.
Figure 13.10 shows the state of the cyber–physical system at the time
of finding an obstacle on the path of the ODG. A MATLAB script that
emulates the operation of the radar and the movement of the ODG is shown
in Figure 13.11.
As soon as the radar detects an obstacle, the data is sent by the controller
to the cloud. The data received from the radar is an array of GPS coordinates
of the obstacle. The computing server calculates a new trajectory, taking
into account the found obstacle, and sends a new route and optimal control
coefficients to the ODG controller. Figure 13.12 shows a new logistics route
after actively responding to an obstacle in the way.
The obstacle tracking algorithm is shown in Figure 13.13.
Figure 13.14 shows the state of the cyber–physical system after obtaining
a new logistics route and optimal control coefficients. ODG follows PGD2 .
 13.5 Simulation Results 347

Figure 13.10 The state of the cyber–physical system at the moment the radar detects an
obstacle (the state of the radar (a), the state of the ODG (b).
348 Modeling of Cyber-Physical Systems

Figure 13.11 MATLAB script responsible for the operation of the radar and movement of
the ODG.

Figure 13.12 New logistics route recalculated due to the occurrence of an obstacle.
 13.5 Simulation Results 349

Figure 13.13 MATLAB script showing the characteristics of the radar and the obstacle
tracking algorithm.
350 Modeling of Cyber-Physical Systems

Figure 13.14 The state of the cyber–physical system after receiving a new logistic route (the
state of the radar (a), the state of the ODG (b).
 13.6 Conclusion 351

13.6 Conclusion
The chapter discusses the most promising in terms of completeness of
description approaches to modeling cyber–physical systems. Not only the
analysis of modeling approaches to cyber–physical systems is carried out but
also the results of building a model for the selected object are presented.
The FCA metamodel approach is excellent for building a high-level
view of the cyber–physical system under development. The result gives a
clear understanding of what systems/functional blocks the cyber–physical
system consists of and how these nodes are interconnected (using cyber or
physical links). Based on the types of incoming and/or outgoing links, you
can determine the type of object. For example, a cyber–physical component
that has no physical inputs and outputs and only interacts in the cybernetic
dimension can be thought of as a computing node. In addition, a component
that receives only physical input and provides only cyber output can be called
a sensor.
The BPMN approach is great for the next design phase—describing
the internal interactions between systems/functional blocks. If the previous
approach showed which components in the system are physical and which are
cyber, then this approach illustrates in detail how these components interact
with each other. What is the nature of the original physical signal from the
steering wheel system to the ODG? Or, what type of cyber signal comes from
the radar to the level of controller interaction. The advantage of this approach
is precisely in detailing the “hidden” interactions between components and
how these interactions form a closed loop of data flow from cyber elements
to physical ones and vice versa.
The last so-called model-based approach, perhaps, in part, is a continu-
ation of the previous one. However, it should be performed most effectively
in conjunction with it. There are many model-based steps worth completing
before starting to use the BPMN-based method. For example, “problem
statement,” “description of physical processes by equations,” “Description of
minimum/maximum parameters and/or restrictions.” Completing these steps
will adequately describe the internal interactions between the components.
The main disadvantage of this approach is the absence of any graphical
representations of intermediate results in the process of building a cyber–
physical system. In turn, its great advantage is algorithmic. Performing each
step, you can get the entire set of data, characteristics, details of technical
implementation for adequate testing of the performance of the developed
cyber–physical system.
352 Modeling of Cyber-Physical Systems

Each approach has its advantages and disadvantages. The authors see the
sense in some symbiosis of these approaches. The first approach is ideal
for the stage of initial prototyping of the problem that has arisen. Follow-
ing the first steps of the third approach “description of physical processes
by equations” and “Description of minimum/maximum parameters and/or
restrictions”, it is possible to collect the necessary set of values/data to
describe the internal interactions between the components of a cyber–physical
system following the second approach. Further, the development of a cyber–
physical system should be continued according to the model-based approach
algorithm.

References
[1] D. Morozov, et al. ‘Multi-paradigm modeling of cyber–physical
Systems’, IFAC-PapersOnLine, 51(11), 2018, pp. 1385–1390. DOI:
https://doi.org/10.1016/j.ifacol.2018.08.334
[2] Graja, et al. ‘BPMN4CPS: A BPMN Extension for Modeling cyber–
physical Systems’, 2016 IEEE 25th International Conference on
Enabling Technologies: Infrastructure for Collaborative Enterprises
(WETICE), 2016, DOI: https://doi.org/10.1109/wetice.2016.41
[3] J. C. Jensen, et al. ‘A model-based design methodology for cyber–
physical systems’, 2011 7th International Wireless Communications and
Mobile Computing Conference, 2011, DOI: https://doi.org/10.1109/iw
cmc.2011.5982785
[4] A. Holovatenko, et al. ‘Energy-Efficient Path-Following Control System
of Automated Guided Vehicles’, J Control Autom Electr Syst 32, 2021,
pp. 390–403. DOI: https://doi.org/10.1007/s40313-020-00668-8
[5] A. Chikrii. Control of Moving Objects in Condition of Conflict. In:
Kuntsevich, V.M. et al. (Eds). Control Systems: Theory and Applica-
tions. Series in Automation, Control and Robotics, River Publishers,
Gistrup, Delft, 2018, pp. 17–43.
[6] V. F. Gubarev, et al. Model Predictive Control for Discrete MIMO
Linear Systems. In: Kondratenko, Y.P., Chikrii, A.A., Gubarev, V.F.,
Kacprzyk, J. (Eds) Advanced Control Techniques in Complex Engi-
neering Systems: Theory and Applications. Dedicated to Prof. V.M.
Kuntsevich. Studies in Systems, Decision and Control, Vol. 203.
Cham: Springer Nature Switzerland AG, 2019, pp. 63-81. DOI:
https://doi.org/10.1007/978-3-030-21927-7_4
 References 353

[7] MathWorks, Automated Driving Toolbox. https://www.mathworks.co

m/products/automated-driving.html, n.d.
[8] Microsoft Azure, Linux Virtual Machines Pricing. https://azure.micros
oft.com/en-us/pricing/details/virtual-machines/linux, n.d.
[9] MathWorks, Instrument Control Toolbox. https://www.mathworks.com/
products/instrument.html, n.d.
[10] IEEE: Standard Computer Dictionary, 1990. A Compilation of IEEE
Standard Computer Glossaries. In NY. 610-1990. ISBN: 1559370793.
[11] S. Meyer, et al. Internet of things architecture iot-a project deliverable
d2.2 concepts for modelling iot-aware processes.
[12] S. Meyer, et al. “Internet of things-aware process modeling: Integrating
iot devices as business process resources,” in 25th International Confer-
ence on Advanced Information Systems Engineering, 2013, pp. 84–98.
[13] S. Meyer, et al., “The things of the internet of things in BPMN,”
in Advanced Information Systems Engineering Workshops, 2015,
pp. 285–297.
[14] K. Sperner, et al., “Introducing entity-based concepts to business process
modeling,” in Business Process Model and Notation - Third Interna-
tional Workshop, BPMN 2011, Lucerne, Switzerland, November 21-22,
2011. Proceedings, 2011, pp. 166–171.
[15] C. Brooks, et al., “Model Engineering Using Multimodeling”, in 1st
International Workshop on Model Co-Evolution and Consistency Man-
agement (MCCM 08), September 2008.
[16] K. Balasubramanian, et al., “Developing Applications Using Model-
Driven Design Environments,” IEEE Computer, vol. 39, no. 2, pp. 33,
February 2006.
[17] G. Karsai, et al., “Model-Integrated Development of Embedded Soft-
ware,” Proceedings of the IEEE, vol. 91, no. 1, January, 2003.
[18] J. Varela-Aldás, et al. (2021) Application for the Cooperative Control of
Mobile Robots with Energy Optimization. In: Kurosu M. (eds) Human-
Computer Interaction. Interaction Techniques and Novel Applications.
HCII 2021. Lecture Notes in Computer Science, vol. 12763. Springer,
Cham. DOI: https://doi.org/10.1007/978-3-030-78465-2_25.
 14
Reliability Control of Technical Systems
based on Canonical Decomposition
of Random Sequences

I. Atamanyuk1,2 , Y. Kondratenko3 , and M. Solesvik4

1 Warsaw University of Life Sciences, Nowoursynowska 166,
02-787 Warsaw, Poland
2 Mykolayiv National Agrarian University, Georgy Gongadze street,

Mykolaiv, 9, 54020, Ukraine

3 Petro Mohyla Black Sea National University, 68 Desantnykiv street,

10, Mykolaiv, 54003, Ukraine

4 Western Norway University of Applied Sciences, Inndalsveien 28,

5063 Bergen, Norway

E-mail: [email protected]; [email protected];
[email protected]

Abstract
A method for determining the suitability of a technical object for further
operation based on the information about the current state and the history
of its functioning during an arbitrary period of time is obtained. The check
is carried out on the basis of the analysis of the state of the object under
study at future points in time and the value of the posterior probability
of no-failure operation. The method for predicting the individual reliability
of objects is based on the canonical decomposition of a random sequence
describing the change in the value of the controlled parameter over time. As
an estimate of the future state, the conditional mathematical expectation is
used; to estimate the probability of no-failure operation, statistical modeling
of a random sequence beyond the observation area is performed.

355
356 Reliability Control of Technical Systems based on Canonical

The method can also be used to control the reliability of the objects
that are characterized by many parameters (in this case, the vector canon-
ical decomposition is used). The proposed approach does not impose any
restrictions on the random sequence of change of the controlled parameters
and, thus, makes it possible to fully take into account the peculiarities of the
functioning of the object under study.
The work presents block diagrams that characterize the features of using
the proposed method. Expressions for assessing the quality of predicting the
state of the controlled object are obtained.

Keywords: Random sequences, canonical decomposition, reliability control,

extrapolation of random sequences.

14.1 Introduction
The main features of the development of modern technology [1] are increas-
ing the degree of automation [2, 3]; increasing loads [4, 5], speeds, tem-
peratures, pressures; reduction in size and weight; increasing requirements
for accuracy and efficiency of functioning [6, 7]; combining individual units
into systems with a single control [8, 9], etc. The increase in complexity
and requirements for performance indicators inevitably leads to the need
to increase requirements for the reliability and durability of equipment
(especially of critical infrastructure systems [10–12]).
The problem of individual prediction [13–16] of state and reliability is of
particular interest. Its solution allows not only to obtain an estimate of the
reliability of each specific sample of products, but also, in the presence of
developed diagnostic support, to move from maintenance by time or resource
to planning operation based on the actual state.
Technical objects are operated under conditions of continuous influence
of different kinds of disturbances, the impact of which has random nature
from the point of view of the intensity, duration and moment of occurrence.
Accordingly, the resulting changes in the state of the operated objects also
turn out to be random [17–19] and form a certain random process in time.
Thus, the problem of predicting the future state of an object from the known
past and present is a problem of a probabilistic-statistical nature, and its
solution is always associated with the study of random processes occurring in
the devices, the state of which is to be predicted. Such studies are constantly
carried out in practice, both in the process of testing products for reliability,
and directly in the process of their operation. This information constitutes
 14.3 Solution 357

the necessary information basis of reliability control of complex technical

objects.

14.2 Problem Statement

Let us introduce the parameter S as a state of the technical control object and
the random sequence {S} = S (i) , i = 1, I as the parameter S changes in a
discrete diapason of time points ti , i = 1, I. The values S (i) , i = 1, I of
the parameter S must satisfy the condition

S (i) ∈ [c, d] , i = 1, I. (14.1)

If the parameter S crosses the boundaries of the admissible region [c, d], a
refusal is recorded. Based on the measurements s (μ) , μ = 1, k, k < I at
discrete moments in time tμ , μ = 1, k, it is required to draw a conclusion
about the suitability of an object for operation at future moments in time
ti , i = k + 1, I.

14.3 Solution
14.3.1 Forecasting the State of Control Objects
The most general approach for solving the problem of assessing the future
state of the control object is based on the Kolmogorov-Gabor polynomial
[20–22]. However, finding its parameters for a large number of known
values and the used order of nonlinear relation is a very complicated and
time-consuming procedure (for example, for 11 known values and order of
nonlinearity 4, it is necessary to obtain and solve 1819 partial differential
equations of the mean square of the extrapolation error). In this regard,
when forming the forecast algorithms realized in practice, various simpli-
fications and restrictions on the properties of a random sequence are used.
For example, Pugachev [23, 24] proposed a number of suboptimal methods
of nonlinear extrapolation with a limited order of stochastic relation based
on the approximation of the posterior probability density of the estimated
vector by the Hermite orthogonal polynomial expansion or in the form of
an Edgeworth series. Solution of the non-stationary equation of Kolmogorov
[25] (a particular case of Stratanovich [26] differential equation for describing
Markov processes) is obtained under the conditions that (a) the diffusion
coefficient is a constant value, and (b) the drift coefficient can be presented as
linear dependence from the state. There are exhaustive solutions of optimal
358 Reliability Control of Technical Systems based on Canonical

linear extrapolation problem for different types of random sequences and dif-
ferent levels of data support in considered forecasting task. Among examples
are (a) Kolmogorov’s equation for random sequences with stationary char-
acter which are measured without errors; Kalman’s technique [27] for noisy
random sequences with Markov properties; Wiener-Hopf’s filter-extrapolator
which is efficient for noisy sequences with stationary character; Kudritskiy’s
algorithms of optimal linear extrapolation [28–30] synthesized using the
Pugachev’s canonical expansion, and others. Nevertheless, we should remem-
ber that only for Gaussian random sequences it is possible to achieve the
highest forecast accuracy using methods of linear extrapolation.
The universal mathematical model (taking into account the restrictions,
which are imposed on the random sequence under consideration) is the
canonical decomposition of such type [31–33]:

!
i !
N
S (i) = M [S (i)] + L(λ)
ν ρ (λ, ν; 1, i) , i = 1, I, (14.2)
ν=1 λ=1

(λ)
where the elements of the canonical expansion Lν , ρ (λ, ν; h, i) are deter-
mined by the recurrent relations:
(λ) 8 9  N (j)
Lν = S λ (ν) − M S λ (ν) − ν−1 μ=1 j=1 Lμ ρ (j, μ; λ, ν) −
 (14.3)
(j)
− λ−1
j=1 νL ρ (j, ν; λ, ν) , λ=1, N , ν = 1, I;

' 9*(
(λ) ) 8
M Lν S h (i) − M S h (i)
(λ)
ρ (λ, ν; h, i) βhν (i) =   
(λ) 2
M Lν
1 ' ( ' ( ' (
= {M S λ (ν) S h (i) − M S λ (ν) M S h (i) −
Dλ (ν)
!
ν−1 !
N !
λ−1
− Dj (μ) ρ (j, μ; λ, ν) ρ (j, μ; h, i) − Dj (ν) ρ
μ=1 j=1 j=1

(j, ν; λ, ν) ρ (j, ν; h, i)}, λ = 1, h, ν = 1, i, h = 1, N , i = 1, I; (14.4)

 14.3 Solution 359
 2  ' ( ' (
Dλ (ν) = M L(λ)
ν =M S 2λ
(ν) − M 2
S (ν) −
λ

!
ν−1 !
N
2
!
λ−1
− Dj (μ) {ρ (j, μ; λ, ν)} − Dj (ν) {ρ (j, ν; λ, ν)}2 ,
μ=1 j=1 j=1

λ = 1, N , ν = 1, I. (14.5)

The coordinate functions ρ (λ, ν; h, i) , ν = 1, i; λ, h = 1, N ; i = 1, I

can be presented as

1, if (h = λ) ∧ (ν = i) ;
ρ (λ, ν; h, i) =
0, if (i < ν) ∨ ((h < λ) ∧ (ν = i)) .

The block-diagram of the calculation algorithm for the parameters

Dλ (ν) , λ=1, N , ν = 1, I and ρ (λ, ν; h, i) , ν = 1, i, λ = 1, N , h =
1, N , i = 1, I of the canonical expansion (14.2) is shown in Figure 14.1.
Nonlinear model (14.2) of the random sequence {S} = S (i) , i =1, I
6 7 (λ)
includes N arrays L(λ) , λ = 1, N , where Li , λ=1, N , i = 1, I are
centered uncorrelated random coefficients. Such coefficients contains data
concerning the corresponding components S λ (i), λ = 1, N , i = 1, I.
The coordinate functions ρ (λ, ν; h, i) , ν = 1, i, λ, h = 1, N , i = 1, I
correspond)to probabilistic
* relations of the order λ + h between the sections
tν and ti , ν, i = 1, I . The expression (14.2) is also true if some stochastic
relations of the random sequence {S} = S (i) , i =1, I are absent. In this
case, the corresponding coordinate functions take the values zero and the data
of the relation are automatically excluded from the canonical expansion.
Suppose that as a result of the measurement, the first value s(1) of
the sequence {S} at the point t1 is known. Therefore, the values of the
coefficients Lλ1 , λ = 1, N are known:

' ( !
λ−1
l1λ = s (1) − M S (1) −
λ λ
l1j ρ(j, λ; 1, 1), λ = 1, I. (14.6)
j=1
360 Reliability Control of Technical Systems based on Canonical

Figure 14.1 Block-diagram of the calculation algorithm for the parameters

Dλ (ν), ρ(λ, ν; h, i) of the polynomial power of canonical expansion (14.3)
 14.3 Solution 361
(1)
Substitution of l1 into (14.2) allows to obtain a polynomial canonical
expansion of the posterior random sequence {S (1,1) } = S(i/S1 (1)):

S (1,1) (i) = S(i/s(1)) = M [S(i)] + (S(1) − M [S(1)])ρ(1, 1; 1, i, )+

!
N
(λ)
!
i !
N
+ L1 ρ(λ, 1; 1, i) + L(λ)
ν ρ(λ, ν, 1, i), i = 1, I. (14.7)
λ=2 ν=1 λ=1

Applying the operation of mathematical expectation to (14.7) gives an

estimate of the future values of the sequence that is optimal by the criterion
of the minimum of the mean-square error of extrapolation, provided that one
value is used to determine this estimate:

m(1,1)
x (1, i) = M [S(i/s(1))] = M [S(i)] + (s(1) − M [S(1)])
ρ(1, 1; 1, i), i = 1, I. (14.8)
Taking into account that the coordinate functions ρ(λ, ν, h, i), λ, h =
1, N , ν, i = 1, I are determined from the condition of the minimum of
the mean-square error of approximation in the intervals between arbitrary
values S λ (ν) and S h (i), the expression (14.8) can be generalized to the case
of forecasting S h (i), h = 1, N , i = 2, I:

m(1,1)
x (h, i) = M [S h (i/x(1))] = M [S h (i)]+
+ (s(1) − M [S(1)])ρ(1, 1; h, i), i = 1, I, (14.9)
(1,1)
where ms (h, i) is the optimal estimate of the future value S h (i), provided
that the value S(1) is used for the forecast.
(2)
Concretization of the second value l1 in (14.7) gives the canonical
expansion of the posterior sequence {S (1,2) } = S(i/s1 (1), s1 (1))2

S (1,2) (i) = S(i/s(1), s(1)2 ) = M [S(i)] + (s(i) − M [s(1)])ρ(1, 1; 1, i)+

+ [s2 (1) − (s(1) − M [S(1)])ρ(1, 1; 2, 1)]ρ(2, 1; 1, i)
!
N
(λ)
!
i !
N
+ Li ρ(λ, 1; 1, i) + L(λ)
ν ρ(λ, ν, 1, i), i = 1, I.
λ=3 ν=2 λ=1
(14.10)
Application of the mathematical expectation operation to (14.10) allows
to obtain an extrapolation algorithm by two values s(1), s(1)2 using the
362 Reliability Control of Technical Systems based on Canonical

expression (14.9):
m(1,2)
s (h, i) = M [S h (i/s(1), s(1)2 )] =
= m(1,1)
s (h, i) + [s2 (1) − m(1,1)
s (2, 1)]ρ(2, 1; h, i), i = 1, I.
(14.11)

The component S N (1) should be used as a posteriori data (for the first
section t1 in the last recurrent cycle)
m(1,N
s
)
(h, i) = M [S h (i/s(1), s(1)2 . . . s(1)N )] =
−1) −1)
= m(i,N
s (h, i) + [sN (1) − m(1,N
s ]ρ(N, 1; h, i), i = 1, I.
(14.12)
and then the forecast is specified by applying the value s(2)

m(2,1)
s (h, i) = M [S h (i/s(1), s(1)2 . . . s(1)N , s(2))] =
= m(1,N
s
)
(h, i) + [s(2) − m(1,N
x
)
(1, 2)]
ρ(1, 2; h, i), i = 1, I. (14.13)
The last recurrent cycle can be presented as the next calculation operation
(for the section t2 ),
ms(2,N ) (h, i) = M [S h (i/s(1), . . . , s(1)N , s(2), . . . , s(2)N )] =
−1) −1)
= m(2,N
s (h, i) + [sN (2) − m(2,N
s (N, 2)]
ρ(N, 2; h, i), i = 1, I.
Generalization of the obtained regularity allows writing a forecast proce-
dure for an arbitrary number of known values [34–36]:
⎧
⎪
⎪ M [S h (i)] for μ = 0;
⎪
⎪
⎪
⎪ (μ,l−1) (μ,l−1)
⎪
⎨ms (h, 1) + (sl (μ) − ms (l, μ))ρ(l, μ; h, i)
(μ,l)
ms (h, i) = for l = 1;
⎪
⎪
⎪
⎪
⎪
⎪ ms
(μ,l−1)
(h, 1) + (sl (μ) − ms
(μ−l,N )
(l, μ))ρ(l, μ; h, i)
⎪
⎩
if l = 1.
(14.14)
(μ,l)
The expression ms (h, i) = M [S h (i)/sν (j), j = l, μ − l, ν =
l, N ; sν (μ), ν = 1, l] for h = l, l = N, μ = k is the optimal estimate
 14.3 Solution 363

m(k,N ) (1, i) of the future value s(i), i = k+, 1, I, provided that the values
sν (j), ν = l, N , j = 1, k are used to calculate this estimate, i.e. the results of
measurements of the sequence {S} at points tj , j = 1, k are known.
(k,N )
The expression for the desired estimate ms (1, i) can be written in the
following explicit form:
!
k !
N
(kN )
m(k,N
s
)
(1, i) = M [S(i)] + (sν (j) − M [S ν (j)])P((j−1)N +ν)
j=1 ν=1
((i − 1)N + 1), (14.15)
where

(α−1) (α−1)
(α) Pλ (ξ) − Pλ (α)γk (i), if λ ≤ α − 1
Pλ (ξ) = (14.16)
γα (ξ), for λ = α


ρ(modN (α), [α/N ] + 1; 1, [α/N ] + 1), for ξ ≤ kN ;
γα (ξ) =
ρ(modN (α), [α/N ] + 1; 1, i), if ξ = (i − 1)N + 1.
(14.17)
[.] is a rounding operation in the expression (14.17).
If the first k values are fixed, then the canonical expansion of a posterior
random sequence {S} can be written in the following way

S (k,N ) (i) = S(i/sν (j), ν = 1, N , j = 1, k) = m(k,N ) (i, j)+

!
i !
N
+ L(λ)
μ ρ(λ, μ; 1, i), i = k + 1, I. (14.18)
μ=k+1 λ=1

The single extrapolation error (for known values sν (j), ν = 1, N , j =

1, k) can be calculated in such way
δ[i/sν (j), ν = 1, N , j = 1, k] = m(k,N
s
)
(1, i) − s(k,N ) (i), i − k + 1, I,
(14.19)

where s(k,N ) (i), i − k + 1, I are the future values of the extrapolated

realization.
The realization s(k,N ) (i), i − k + 1, I in the area of forecasting develops
in a random way, so its exact values cannot be specified. However, the expres-
sion (14.18) is an exact model of the stochastic evolution of the realization
364 Reliability Control of Technical Systems based on Canonical

s(k,N ) (i), i − k + 1, I, which allows to reduce the expression (14.19) to the

form

Δ[i/sν (j), ν = 1, N , j = 1, k] = m(k,N

x
)
(i, j) − S (k,N ) (i), i = k + 1, I.
(14.20)

Thus, the error of a single extrapolation has a stochastic nature that deals
with the random character of the investigated processes.
Implementation of the mathematical expectation procedure to (14.20),
taking into account the expression (14.18), give us possibility to calculate
the systematic component of the single extrapolation’s error

ε[i/sν (j), ν = 1, N , j = 1, k] = M [Δ(i/sν (j), ν = 1, N , j = 1, k)] =

!
i !
N
= M [L(λ)
μ ]ρ(λ, μ, 1, i) = 0, i = k + 1, I. (14.21)
μ=k=1 λ=1

The latter Equation (14.21) illustrates that the procedures (14.14), (14.15)
provides calculation (for each predicted realization) of unbiased estimates of
future values.
This also underlines that the proposed algorithm gives possibility to
provide the unbiasedness for the average

ε(k,N ) (i) = M [ε(i/sν (j), ν = 1, N , j = 1, k)] = 0, i = k + 1, I. (14.22)

The variance help us to estimate the significance of the random compo-

nent of the error for single extrapolation

DΔ [i/sν (j), ν = 1, N , j = 1, k] = M [(Δ[i/sν (j), ν = 1, N , j = 1, k]−

1 l 22 
! ! N
− ε[i/χ (j), ν = 1, N , j = 1, k]) ] = M
ν 2 (λ)
Lμ ρ(λ, μ; 1, i)
μ=k+1 λ=1

!
i !
N
= Dλ (μ){ρ(λ, μ; 1, i)}2 , i = k + 1, I. (14.23)
μ=k+1 λ=1

Since the specific values of the extrapolated realization do not impact

to the variance DΔ [i/sν (j), ν = 1, N , j = 1, k] it is possible to form an
 14.3 Solution 365

expression for the variance of the posterior random sequence

(k,N )
DΔ (i) = M [DΔ [i/X ν (j), ν = 1, N , j = 1, k]]
!
i !
N
(λ)
= Dλ (μ)ρ(λ, μ; 1, i)β1μ (i). (14.24)
μ=k+1 λ=1

Taking into account (14.22), (14.24), the expression for the mean square
of the forecast error can be written as
(k,N )
E (k,N ) (i) = [ε(k,N ) (i)]2 + DΔ (i) =

(k,N )
!
i !
N
DΔ (i) = Dλ (μ)ρ(λ, μ; 1, i), i = k + 1, I. (14.25)
μ=k+1 λ=1

Thus, the mean square of the forecast error based on the procedures
(14.14), (14.15) is equal to the variance of the posterior random sequence.
It is necessary to note that the parameters of the procedure (14.14),
calculated by the size of the statistical sample of realizations, can be specified
using new a posteriori information sL+1 (ν), ν = 1, I in real time in the
process of solving the forecasting problem:

L+1 
L
sni sni L
n l=1 l=1snL+1 L M(L+1) [S n ]L + snL+1
M(L+1) [S ] = = = = ;
L+1 (L + 1)L L+1 (L + 1)
 (n) 2
L+1 L
(n)
(lν,1 ) (lν,1 )2 (L − 1) (n)
l=1 l=1 (lν,1 )2
Dn(L+1) (ν) = = +
L L(L − 1) (L)
n
Dn(L+1) (ν) + (S(L+1) (ν) − m(n−1,ν) (n, ν))
;
L

L+1
(n)
lν,l (Slλ − M(L+1) [S λ (i)])
l=1
ρL+1 (n, ν; λ, i) =
LDn,(L+1) (ν)
ρL+1 (n, ν; λ, i)(L − 1) + (snL+1 (ν) − m(n−1,ν) (n, ν))
(sλL+1 (i) − M(L+1) [S λ (i)])
=
LDn,(L+1) (ν)
Thus, the results of extrapolation can be used for significant (up to
several arithmetic operations) simplification of the recalculating procedure
366 Reliability Control of Technical Systems based on Canonical

for the algorithm (14.14) parameters in the case if new statistical data of the
investigated random sequence come. It allows to use the updated parameters
for the current solution of the forecast problem.
By analogy with the procedure of forming the algorithm (14.14), the
method of extrapolation of the realization of the vector random sequence {S̄},
taking into account the stochastic relations M [Sjl (μ)Shq ], μ, i = 1, I; j, h =
1, H; l, q = 1, N between the components, can be obtained as following:
⎧
⎪
⎪ M [Sh (i)], if μ = 0;
⎪
⎪
⎪
⎪
⎪
⎪
(μ,l−1) (μ,l−1) (h,q)
ms;j,h (q, i) + (slj (μ) − mx;j,h (l, μ))βj,i (μ, i),
⎪
⎪
⎪
⎪
⎪
⎨
for l > 1, j > 1;
(μ,l)
ms;j,h (q, i) = m(μ,N ) (q, i) + (sl (μ) − m(μ,N ) (l, μ))β (h,q) (μ, i),
⎪
⎪ s;j−1,h j x;j−1,j j−1,N
⎪
⎪
⎪
⎪ for l = 1, j > 1;
⎪
⎪
⎪
⎪ (μ−1,N ) (μ−1,N ) (h,s)
⎪
⎪ m (q, i)+(sl1 (μ) − ms;H,1 (l, μ))β1,1 (μ + 1, i),
⎪ s;H,h
⎩
if l = 1, j = 1;
(14.26)
(μ,l)
where ms;j,h (1, i) = M [Sh (i) /snλ (ν) , λ = 1, H, n = 1, N , ν =
1, μ − 1; snλ (μ) , λ = 1, j] is the optimal, in the mean-square sense, estimate
of the future values of the investigated random sequence, provided that a
posteriori information snλ (ν), λ = 1, H, n = 1, N is used for the forecast.
The algorithm is based on the canonical model

!
i−1 !
H !
N
(λ) (h,1)
Sh (i) = M [Sh (i)] + Lνl βlλ (ν, i)
ν=1 l=1 λ=1
!
h−1 !
N
(λ) (h,1) (1)
+ Lil βlλ (i, i) + Wih , l = 1, I; (14.27)
l=1 λ=1

(λ)
!
ν=1 !
H !
N
(l,λ)
Lνl = Slλ (ν) − M [Slλ (ν)] − L(l)
μm βmj (μ, ν)−
μ=1 m=1 j=1

!
l−1 !
N !
λ−1
(j) (l,λ)
− νm (ν, ν) −
L(j) Lνl βlj (ν, ν), ν = 1, I; (14.28)
m=1 j=1 j=1
 14.3 Solution 367
   ' ( ' (
(λ) 2
Dl,λ (v) = M Lvl = M Sl (v) − M Sl (v)
2λ 2 λ

!
v−1 !
H !
N 2 !
l−1 !
N 2
(l,λ) (l,λ)
− Dmj (μ) βmj (μ, v) − Dmj (v) βmj (v, v)
μ=1 m=1 j=1 m=1 j=1

!
λ−1 2
(l,λ)
− Dlj (v) βlj (v, v) , v = 1, I; (14.29)
j=1

'  ' ((
(λ) (λ)
M Lvl Shq (i) − M Lvl Shq (i)
(h,q)
βl,λ (v, i) =   
(λ) 2
M Wvl
1 ' ( ' ( 8 9
= M Slλ (v)Shq (i) − M Slλ (v) M Shq (i)
Dl,λ (v)
!
v−1 !
H !
N
(l,λ) (h,s)
− Dmj (μ)βmj (μ, v)βmj (μ, i)
μ=1 m=1 j=1

!
l−1 !
N
(l,λ) (h,s)
− Dmj (v)βmj (v, v)βmj (v, i)
m=1 j=1

!
λ−1
(l,λ) (h,s)
− Dlj (v)βmj (v, v)βmj (v, i), λ = 1, hv = 1, i. (14.30)
j=1

The difference with (14.14) lies in that that after taking full account of the
posterior information sj (μ), s2j (μ)...sN
j (μ) at the fixed section tμ , the transi-
tion to the accumulation is carried out to predict the posterior information of
the next component Sj+1 , in the case j=N the transition to the next section
tμ+1 is carried out. Diagram in Figure 14.2 reflects the peculiarities of the
computational process when using the extrapolator (14.26).
The expression for the mean-square error of extrapolation using the
algorithm (14.26) by the known values snj (μ), μ = 1, k; j = 1, H; n = 1, N
has the form
(k,N ) 8 9
Eh (i) = M Sh2 (i) − M 2 [Sh (i)] −
!
k ! H ! N 2
(h,1)
− Djn (μ) βjn (μ, i) , i = k + 1, I. (14.31)
μ=1 j=1 n=1
368 Reliability Control of Technical Systems based on Canonical

Figure 14.2 Generative procedure for future values of a random sequence based on the
algorithm (14.26)

(K,N )
The mean-square error of extrapolation Eh (i) is equal to the variance
of the posterior random sequence
(k,N ) ) * (k,N )
Sh (i) = S i/sνl (j) , ν = 1, N , j = 1, k, l = 1, H = mH,h (1, i)

!
i−1 !
H !
N
(λ) (h,1)
!
h−1 !
N
(λ) (h,1) (1)
+ Lνl βlλ (ν, i) + Lil βlλ (i, i) + Lih ,
ν=k+1 l=1 λ=1 l=1 λ=1

i = k + 1, I. (14.32)

If the vector random sequence has only linear stochastic relations

(M [Sj (μ)Sh (i)] = 0, .M [Sjl (μ)Shs (i)] = 0, l + s >) then the extrapolation
algorithm takes the form [37–39]:
⎧
⎪
⎪ M [Xh (i)] , if μ = 0;
⎪
⎪
⎪
⎪ ' (
⎨ (μ,l−1) (μ,l−1) (l)
mx;h (i) = mx;h (i) + xl (μ) − mx;h (μ) ϕhμ (i), if l = 1;
(μ,l)
⎪
⎪
⎪
⎪ ' (
⎪
⎪
⎩m(μ,H) (i) + x1 (μ) − m(μ−1,H) (μ) ϕ(1) (i), if l = 1;
x,h x;h hμ
(14.33)
 14.3 Solution 369
(μ,l)
where mx;h (i) = M [Xh (i)/Xλ (v ), λ = 1, H,v = 1, μ − 1; xj (μ),
j = 1, l]h = 1, H,i = k, I is the linear optimal by the criterion of the
minimum of a mean-square error of forecasting estimate of the future values
of the investigated sequence, provided that the values sλ (v), λ = 1, H,v =
1, μ − 1; sj (μ), j = 1, l are known.
The parameters of the algorithm are the elements of the linear vector
canonical decomposition

!
i !
H
(λ)
Sh (i) = M [Sh (i)] + vv(λ) ϕhv (i), i = 1, I, (14.34)
v=1 λ=1

where
!
v−i !
H
(j)
Vv(λ) = Sλ (v) − M [Sλ (v)] − vμ(j) ϕλμ (v)
μ=1 j=1

!
λ−1
(j)
− Vv(j) ϕλμ (v), v = 1, I; (14.35)
j=1

(λ)
(λ) M [Vv (Sh (i) − M [Sh (i)])] 1
ϕhv (i) =   = (M [Sλ (v)Sh (i)]
(λ) 2 Dλ(v)
M {Vv }

!
v−1 !
H
(j) (j)
− M [Sλ (v)]M [Sh (i)] − Dj (μ)ϕλμ (v)ϕhμ (i)
μ=1 j=1

!
λ−1
(j) (j)
− Dj ϕλv (v)ϕhv (i), λ = 1, I, v = 1, i ; (14.36)
j=1

 2  ' (
Dλ (v) = M Vv(λ) = M {Xλ (v)}2 − M 2 [Xλ (v)]

!
v−1 !
H 2 !
λ−1 2
(j) (j)
− Dj (μ) ϕλμ (v) − Dj (v) ϕλv (v) , v = 1, I.
μ=1 j=1 j=1
(14.37)
370 Reliability Control of Technical Systems based on Canonical

The expression for the mean-square error of extrapolation using the

algorithm (14.33) is written as

!!
μ−1 H 2 !
l−1 2
(μ,l) (j) (j)
Eh (i) = Dh (i) − Dj (v) ϕhv (i) − Dj (μ) ϕhμ (i) ,
v=1 j=1 j=1

i = μ + 1, I. (14.38)

14.3.2 Identification of Random Sequences Model’s Parameters

based on Statistical Goodness-of-fit Tests
The coordinate functions allow taking into account (a) the mechanism of after
effect and (b) estimation of the stochastic relation in the canonical expansion
(14.2):
(λ) (λ) o
M [Lv (S(i) − M [S(i)])] M [Lv X (i)]
ρ(λ, v; 1, i) = = , i > v.
Dλ (v) Dλ (v)
(14.39)
and these functions determine the level (degree) of stochastic relationship
between subsequent sections of the random sequence S (i) , i > ν and
(λ) (λ)
coefficients Lν . Thus, we can assume that the coefficients Lν in the section
tν have zero’s influence to the subsequent values of the random sequence,
if (starting from any ikν > ν), ρ (λ, ν; 1, i) ≡ 0 is true for the arbitrary
valueλ. Accordingly, the after effect interval kν from the sampling point tν is
determined in this case as
kv = ikv − v. (14.40)
In the situation under consideration, it is possible to form the components
of the canonical expansion using finite sample of the size L. Thus, the
process of kv determination can be transformed to solving the next prob-
lem: 1) according to the sample data, estimates of the coordinate functions
ρL (λ, v; 1, i), λ = 1, N , i = v, I can be obtained; (b) it is necessary to find
such value ikv , v < ik for which the identity ρ(λ, v; 1, i) ≡ 0, i > v is
satisfied with a set degree of confidence.
Analyzing the model (14.2) and its linearization mechanism it is possible
to transform the problem of after effect estimation and determination of the
nonlinearity order of a random sequence to the problem of the linear relation
(λ)
between Lv and S(i), v < i, the standard quantitative characteristic of
 14.3 Solution 371

which is the normalized correlation coefficient

' (
(λ)
M Lv S(i)
r(λ) (v, i) = < < , λ = 1, N , v = 1, I, i = v, I. (14.41)
Dλ (v) Dx (i)
The expression (14.41), using (14.39), can be transformed to the follow-
ing form <
Dλ (v)ρ(λ, v; 1, i)
(λ)
r (v, i) = < , i > v, (14.42)
Dx (i)
where the components Dλ (v), ρ(λ, v; 1, i) of the canonical expansion are
used for calculation of the correlation coefficient r(λ) (v, i). Since the esti-
mates of all these components can be obtained during the initial statistical
data processing, it is possible to estimate the normalized correlation coeffi-
cient r(λ)(L) (v, i) for any values v,λ and i. From this point of view, we can
reformulate the task of significance estimation for the correlation between the
(λ)
coefficient Lv and the ith section of the investigated random sequence to the
problem of testing the statistical hypothesis

r(λ) (v, i) = 0 (14.43)

and the alternative hypothesis r(λ) (v, i) = 0.

The random variable
 
1 1 + r(λ)(L) (v, i)
ln
2 1 − r(λ)(L) (v, i)

should be considered distributed normally, with the mathematical expecta-

tion [40]  
1 1 + r(λ)(L) (v, i) r(λ) (v, i)
m = ln +
2 1 − r(λ)(L) (v, i) 2(L − 1)
1
and the variance D = L−3 ,
therefore the value
  
√ 1 1 + r(λ)(L) (v, i)
a (v, i) = L − 3 ln
(λ)
2 1 − r(λ)(L) (v, i)
1   2
1 1 + r(λ)(L) (v, i) r(λ) (v, i)
− ln +
2 1 − r(λ)(L) (v, i) 2(L − 1)
372 Reliability Control of Technical Systems based on Canonical

has the standard normal distribution (0,1).

Thus, the data can be consistent with the hypothetical value r(λ) (v, i)
(with the significance level), if the value
 
1 1 + r(λ)(L) (v, i) r(λ) (v, i)
ln +
2 1 − r(λ)(L) (v, i) 2(L − 1)

lies within
     
1 1 + r(λ)(L) (v, i) zα 1 1 + r(λ)(L) (v, i) zα
ln −√ , ln +√
2 1 − r(λ)(L) (v, i) L−3 2 1 − r(λ)(L) (v, i) L−3
where zα is the value of the standard normal deviate which corresponds to the
level of confidence α. In other cases, the hypothetical value r(λ) (v, i) does not
correspond to the statistics.
Because the condition (r(λ) (v, i) = 0) is checking, the random variable
(λ)
a (v, i) takes the form [41, 42]
 
√ 1 + r (λ)(L) (v, i)
a(λ) (v, i) = L − 3 ln
1 − r(λ)(L) (v, i)

and the hypothesis (14.43) can be accepted if the following condition is

fulfilled
− zα < a(λ) (v, i) < zα . (14.44)
Hypothesis (14.44) is multi-tested many by increasing the parameter λ
until the certain boundary value N (v,i) at which condition (14.44) is true
(N (v,i) is a higher order of the nonlinear relation between the sections tv
and) tv After that, the number of interval increases (i = i + 1) and the search
procedure for the higher order of nonlinearity is repeated for a new interval.
If for some ikv of the area i > v and the arbitrary λ (as a rule, it is
sufficient to check at λ=1 : nonlinear relations decay faster than linear ones)
the statement r(λ) (v, i) = 0,i > ikv turns out to be true, then this means that
the after effect interval for the sampling point tv is equal to ikv − 1 and for all
i > ikv the value of the coordinate function ρ(λ, v; 1, i) should be taken to be
zero.
By checking the after effect for all sampling points tv , v = 1, I ∗ (in
which the random sequence S is investigated) it is possible to determine the
parameters N and I of the canonical expansion (14.2) as
I = max (ikv − v) ,
v
 14.3 Solution 373

N = max N (v,i) .
v,i

14.3.3 Method of Reliability Control of Technical Objects

Taking into account that the values of the controlled parameter S change in
the forecast area in a random manner, the comprehensive characteristic of the
reliability of the investigated technical object is the probability of failure-free
operation

P (k) (I) = P c < S (k) (i) < d, i = k + 1, I/x(μ), μ = 1, k . (14.45)

Thus, the problem is reduced to determining the probability

) that the
* real-
ization of the a posteriori random sequence S (k) i/s(μ), μ = 1, k , i =
k + 1, I will not go beyond the boundaries of the admissible region [A; d].
The probability estimate P (k) (I) can be obtained as a result of a statistical
experiment using the expression
m
P ∗(k) (I) = (14.46)
L
where m – is the number of realizations that satisfy the condition (14.1);
L – the total number of realizations.
Thus, on the basis of the above, the method of the predictive control of the
failure-free operation of technical objects consists of the realization of the
following stages:
• collection of statistical data on the functioning of the same type of
objects (for objects with one parameter and for objects with many
parameters);
• estimation of discretized moment functions;
• calculation of the parameters of the mathematical model (14.2)—
for objects with one parameter; (14.27)—for objects with many
parameters);
• calculation of the state of the control object at future moments in time
(algorithm (14.14) or (14.26));
• verification of the condition (14.1) (in case of non-fulfillment—the
operation of the object is stopped and the subsequent analysis is
stopped);
(λ)
• multiple modeling of the values of random coefficients Li , i =
(λ)
k + 1, I, λ = 1, N or Lνl (for objects with many parameters) with
374 Reliability Control of Technical Systems based on Canonical

a given distribution law and with the subsequent formation, using the
expression (14.18) or (14.32), of a set of possible continuations of the
realization of the investigated random sequence in the forecast area
[tk+1 ...tI ];
• verification of the condition of non-intersection of the boundaries of the
admissible region of change of the controlled parameters by the obtained
trajectories and determination of an estimate of the probability of failure-
free operation of a technical object;
• making a decision on the subsequent operation of the object based on
the obtained estimate of the probability of failure-free operation.

14.4 Conclusion
Thus, a method of reliability control of technical objects, the parameters
of which are of a stochastic nature, is obtained in the work. An individual
reliability estimation based on the information about the functioning of an
object during operation is a peculiarity of the proposed approach. Making
decision on suitability at future points in time is carried out in two stages: at
the first stage, the belonging of the predicted parameters to the permissible
region is checked, and at the second stage, the compliance of the proba-
bility value of no-failure operation with the required level is checked. The
mathematical model based on the canonical decomposition does not imply
any restrictions on the random sequence of change of the parameters of the
control object, which makes it possible to take into account the features of
its functioning as much as possible. Expressions for the mean-square error of
extrapolation are obtained, which allow us to estimate the accuracy of solving
the forecasting problem. The resulting method can be used to control the
reliability of arbitrary technical objects and systems [9, 10, 43–45], as well
as their components.

References
[1] A. Adarsh, R. Mangey, ‘System reliabilty: solutions and technologies’,
Taylor & Francis Group, 2019.
[2] V. Varakuta, T. Hlimancov, S. Starodubcev, S. Ptashka, ‘Determination
of risk indicators of loss of negative situations on technological danger-
ous objects’, Science and Technology of the Air Force of Ukraine, Issue
1(34), 2019, pp. 107–116.
 References 375

[3] P.M. Frank, ‘Fault diagnosis in dynamic systems using analytical and
knowledge-based redundancy: A survey and some new results’, J.
Automatica 3, 459–474, 1990.
[4] R.J. Patton, P.M. Frank, R.N. Clarke., ‘Fault diagnosis in dynamic
systems: theory and application’, Prentice-Hall Inc., 1989.
[5] R.J. Patton, R.N. Clark, P.M. Frank, ‘Issues of fault diagnosis for
dynamic systems’, Springer Science & Business Media, 2000.
[6] P.M. Frank, X.-C. Ding, ‘Survey of robust residual generation and
evaluation methods in observer-based fault detection systems’, Journal
of process control 7.6, 403–424, 1997.
[7] S.X. Ding, ‘A unified approach to the optimization of fault detection
systems’, International journal of adaptive control and signal processing
14.7, 725–745, 2000.
[8] N. Silva., M. Vieira, ‘Towards Making Safety-Critical Systems Safer:
Learning from Mistakes’, In: ISSREW, 2014, 2014 IEEE Interna-
tional Symposium on Software Reliability Engineering Workshops
(ISSREW), 162–167, 2014.
[9] V.M. Kuntsevich, et al (Eds), ‘Control Systems: Theory and Applica-
tions’, Series in Automation, Control and Robotics, River Publishers,
Gistrup, Delft, 2018.
[10] Y.P. Kondratenko, V.M. Kuntsevich, A.A., Chikrii V.F. Gubarev, (Eds),
‘Advanced Control Systems: Theory and Applications’, Series in
Automation, Control and Robotics, River Publishers, Gistrup, 2021.
[11] V. Varakuta, T. Hlimancov, S. Starodubcev, S. Ptashka, ‘Determination
of risk indicators of loss of negative situations on technological danger-
ous objects’, J. Science and Technology of the Air Force of Ukraine, No.
1(34), pp. 107–116. https://doi.org/10.30748/nitps.2019.34.15.
[12] A.V. Boyarchuk, etc., ‘Safety of Critical Infrastructures: Mathematical
Analysis and Engineering Methods of Analysis and Ensuring’, National
Aerospace University named after N.E.Zhukovsky “KhAI”, Kharkiv,
2011.
[13] M. Nafria, ‘Circuit reliability prediction: challenges and solutions for
the device time-dependent variability characterization roadblock’, IEEE
Latin America Electron Devices Conference (LAEDC), Mexico, 19–21
April 2021.
[14] H. Jiang, ‘Time Dependent Variability in Advanced FinFET Technology
for End-of-Lifetime Reliability’, IEEE International Reliability Physics
Symposium (IRPS) Monterey, CA, USA, 21–25 March 2021.
376 Reliability Control of Technical Systems based on Canonical

[15] H. Yu, Y. Ran, G.B. Zhang, G. Ying, ‘A dynamic time-varying relia-

bility model for linear guides considering wear degradation’, Nonlinear
Dynamics, Volume 103, pp. 699–714, 2021.
[16] A. Kremer, L. Dücsö, B. Bertsche, ‘Reliability Prediction using Design
of Experiments’, Proceedings of the 30th European Safety and Reli-
ability Conference and the 15th Probabilistic Safety Assessment and
Management Conference, Venice, Italy, 1–5 November 2020.
[17] A.V. Palagin, V.N. Opanasenko, ‘Design and application of the PLD-
based reconfigurable devices.’ In: Design of Digital Systems and
Devices: Adamski M., Barkalov A., Wegrzyn M. (Eds.), Lecture Notes
in Electrical Engineering. Verlag, Berlin, Heidelberg: Springer, Volume
79, pp. 59–91, 2011.
[18] Sokolov Y.N., Kharchenko V.S., Illyushko V.M., et al, ‘Applications of
Computer Technologies for Software-Hardware Complexes Reliability
and Safety Assessment Systems’, National Aerospace University named
after N.E.Zhukovsky “KhAI”s, Kharkiv, 2013.
[19] Odarushchenko O.N., Ponochovny Y.L., Kharchenko V.S., et al, ‘High
Availability Systems and Technologies’, National Aerospace University
named after N.E.Zhukovsky “KhAI”, Kharkiv, 2013.
[20] P.C. Jack, ‘Statistical techniques in simulation’, Publishing house Klei-
jnen, 1978.
[21] C. Chatfield, ‘Time series forecasting’, London: Chapman, and Hall,
2000.
[22] E. Wentzel, ‘Theory of Probability and its Engineering Applications’,
Nauka, Moscow, 1998.
[23] V. Pugachev, ‘Theory of Random Functions: And Its Application to
Control Problems’, Pergamon Press, London, 2013.
[24] V. Pugachev, I. Sinitsyn, ‘Stochastic systems: theory and applications’,
Word Scientific Publishing, London, 2002.
[25] B.S. Everitt (Ed.), ‘The Cambridge dictionary of statistics’, Cambridge
University press, New York, 2006.
[26] R.S. Tsay, ‘Nonlinear time series models: testing and applications.
Course in Time Series Analysis’, New York: Wiley, 2001.
[27] D.C. Montgomery, G.C. Runger, ‘Applied Statistics and Probability for
Engineers’, Wiley Custom, 2012.
[28] V.D. Kudritskii, ‘Generalized optimal linear extrapolation algorithm
for realizations of random sequences’, Engineering Simulation, 12(3),
pp. 339–348, 1995.
 References 377

[29] V.D. Kudritskii, ‘Optimal linear filtering and extrapolation of realiza-

tions of a vector random function observed with errors’, Cybernetics
and Systems Analysis, 36(4), pp. 579–584, 2000.
[30] V.D. Kudritskii, ‘Algorithm of Wiener filtering and extrapolation
for nonstationary random processes observed with correlated errors’,
Cybernetics and Systems Analysis, 35(3), pp. 406–412, 1999.
[31] I. Atamanyuk, Y. Kondratenko, ‘Canonical mathematical model and
information technology for cardio-vascular diseases diagnostics’, Pro-
ceedings of the 14th International Conference The Experience of
Designing and Application of CAD Systems in Microelectronics,
CADSM 2017, Lviv – Polyana, Ukraine, pp. 438–440 May 2017. DOI:
10.1109/CADSM.2017.7916170
[32] Y. Kondratenko, et al, ‘Robotics and Prosthetics at Cleveland State Uni-
versity: Modern Information, Communication, and Modeling Technolo-
gies.’ Ginige, A. et al. (Eds): Information and Communication Tech-
nologies in Education, Research, and Industrial Applications. ICTERI
2016. Communications in Computer and Information Science, vol. 783.
Springer, Cham, pp. 133–155, 2017. DOI: https://doi.org/10.1007/978-
3-319-69965-3_8
[33] Y. Kondratenko, et al, ‘University Curricula Modification Based on
Advancements in Information and Communication Technologies.’ Pro-
ceedings of the 12th International Conference on Information and
Communication Technologies in Education, Research, and Industrial
Application. Integration, Harmonization and Knowledge Transfer, 21–
24 June, 2016, Kyiv, Ukraine, Ermolayev, V. et al. (Eds), ICTERI’2016,
CEUR-WS, Vol-1614, pp. 184–199, 2016.
[34] V. Shebanin, et al, ‘Application of fuzzy predicates and quantifiers by
matrix presentation in informational resources modeling’, Proceedings
of XII International Conference “MEMSTECH 2016”, 22-24 April
2016, Lviv-Poljana, pp. 146–149.
[35] I.P. Atamanyuk, ‘Algorithm of extrapolation of a nonlinear random
process on the basis of its canonical decomposition’, J. Cybernetics and
Systems Analysis, 41(2), pp. 267–273, 2005.
[36] I.P. Atamanyuk, ‘Optimal Polynomial Extrapolation of Realization of
a Random Process with a Filtration of Measurement Errors’, Jour-
nal of Automation and Information Sciences, Volume 41, Issue 8,
Begell House, USA, pp. 38–48, 2009. DOI: 10.1615/JAutomatInf-
Scien.v41.i8.40
378 Reliability Control of Technical Systems based on Canonical

[37] V. Shebanin, et al., ‘Simulation of vector random sequences based on

polynomial degree canonical decomposition’, Eastern-European Journal
of Enterprise Technologies, Vol. 5, No. 4 (83), 2016, pp. 4–12, DOI:
http://dx.doi.org/10.15587/1729-4061.2016.80786
[38] I. Atamanyuk, Y. Kondratenko, N. Sirenko, ‘Forecasting Economic
Indices of Agricultural Enterprises Based on Vector Polynomial Canon-
ical Expansion of Random Sequences’, Proceedings of the 12-th
Int. Conference ICTERI’2016, CEUR-WS, 21-24 June, 2016, Kyiv,
Ukraine, Vol-1614, pp. 458–468, 2016.
[39] I. Atamanyuk, Y. Kondratenko, N. Sirenko, ‘Management System for
Agricultural Enterprise on the Basis of Its Economic State Fore-
casting’, Complex Systems: Solutions and Challenges in Economics,
Management and Engineering, Vol. 125, pp. 453–470, 2018. DOI:
https://doi.org/10.1007/978-3-319-69989-9_27
[40] E. Lehmann, J. Romano, ‘Testing Statistical Hypotheses’, Springer-
Verlag New York, 2005, p. 786.
[41] C. Walck, ‘Handbook on statistical distributions for experimentalists,’
Particle Phys. Group, Fysikum, Univ. Stockholm, Stockholm, Sweden,
Tech. Rep. SUF-PFY/96-01, 2007.
[42] G.E.P. Box, G.M. Jenkins, ‘Time–series Analysis, Forecasting and
Control’, Holden–Day, San Francisco, 1970.
[43] A.N. Tkachenko, et al. ‘Evolutionary adaptation of control processes
in robots operating in non-stationary environments.’ J. Mechanism and
Machine Theory. – Printed in placecountry-regionGreat Britain, 1983. –
Vol. 18, No. 4. – pp. 275–278. DOI: 10.1016/0094-114X(83)90118-0
[44] I.P. Atamanyuk, et al, ‘The Algorithm of Optimal Polynomial Extrap-
olation of Random Processes.’ In: Modeling and Simulation in Engi-
neering, Economics and Management, K.J. Engemann, et al. (Eds.),
International Conference MS 2012, New Rochelle, NY, USA (May 30
– June 1, 2012), Proceedings. Lecture Notes in Business Information
Processing, Volume 115, Springer, 2012, pp. 78–87. DOI: 10.1007/978-
3-642-30433-0_9
[45] M. Solesvik, et al, ‘Joint Digital Simulation Platforms for Safety and
Preparedness.’ In: Y. Luo (Ed), Cooperative Design, Visualization,
and Engineering. CDVE 2018. Lecture Notes in Computer Science,
vol. 11151. Springer, Cham, pp. 118–125, 2018.
 15
Petunin Ellipsoids in Automatic Control
Systems Design

D.A. Klyushin, S.I. Lyashko, and A.A. Tymoshenko

Taras Shevchenko National University of Kyiv, Prospekt Glushkova 4D,

03680, Kyiv, Ukraine
E-mail: [email protected]; [email protected];
[email protected]

Abstract
Mathematical modeling of modern automatic control systems requires storing
and processing large amounts of data. Complex systems are described by
numerous parameters that have different significance and obey certain restric-
tions. These constraints often take the form of a system of linear inequalities.
This paper proposes a new way of describing the optimization domain with
linear constraints in the design space using Petunin ellipsoids. The statistical
properties of these ellipsoids are given, making them an effective tool for
describing data in the design space. The problem of construction of the
optimal Petunin ellipsoids in design space with linear constrains is consid-
ered in the paper. The algorithms of construction of the Petunin ellipsoids
without constraints and with linear constraints are described. The issues of
the complexity and statistical features are analyzed.
The second part is devoted to a new distribution-free approach to comput-
ing ellipsoidal conformal prediction set based on the Hill’s assumption and
the Petunin ellipsoids. The resulting prediction set is a Petunin ellipsoid with
precise confidence level depending on the number of points only. We show
that our method works for classical distributions (uniform and Gaussian) with
high performance. In addition, our method allows determining the median
of the random points set (the deepest point), outliers (outer points) and the
statistical depth for every point of the set.

379
380 Petunin Ellipsoids in Automatic Control Systems Design

Keywords: Petunin ellipsoids, minimum volume ellipsoid, minimum trace

ellipsoid, constrained optimization, design space, data mining, prediction set,
outlier detection.

15.1 Introduction
Modern computer-aided design of automatic control systems is impossible
without data analysis that describes the parameters of objects and the results
of experiments in the form of multidimensional vectors in the design space.
The complexity of the problem is due to the fact that the number of such
parameters is very large and therefore it is desirable to single out among them
those that are important from a certain point of view (for example, statistically
significant) and discard points that can be interpreted as outliers. Moreover,
each parameter, as a rule, is subject to certain restrictions. It is either limited
to the minimum and maximum values, or is related to other parameters by
linear relationships.
The problems of designing technical systems are often reduced to solving
variational problems of minimizing functionals subject to the fulfillment of
constraints in the form of a system of linear inequalities. This leads to the need
to determine the optimization area in the design space, in which the search
for the optimal solution that satisfies the given constraints is performed. A
large scatter of parameters in the design space can complicate mathematical
modeling and worsen the properties of numerical optimization algorithms,
so the optimization area should be compact and contain only statistically
significant points. In addition, many efficient optimization algorithms make
significant use of the geometric properties of the optimization domain (for
example, convexity). It is desirable that the optimization domain be convex
and described by simple equations. These requirements are met by ellipsoids
covering the given points. Ellisoids of minimum volume are often used as
such ellipsoids, ellipsoids with minimal traces, Dikin ellipsoids and other
modifications. The choice of these ellipsoids is dictated by geometric consid-
erations and does not take into account the statistical nature of the parameters
of the designed technical systems. demands without considering the data
volume.
The statistical nature of the source data dictates the choice of the appro-
priate tool. In particular, similar problems were considered in [1], where a
comparative analysis of several methods for constructing optimal descriptions
of design data was carried out, in particular, an ellipsoid of minimum volume,
an ellipsoid with a minimum trace, and the Monte Carlo method. Despite the
 15.2 Petunin Ellipses and Their Statistical Properties 381

effectiveness of the methods demonstrated, they have a big drawback-they

do not take into account the statistical nature of the parameters, focusing on
geometric properties. The Petunin ellipsoid described in [3–8] allows us to
elegantly solve this problem.
The second topic discussed in the paper is construction of prediction sets,
which was investigated in many papers. They are closely tied with the notions
of the statistical depth, the multivariate ordering and the quintile contour con-
struction. In particular, using B.M. Hill’s work [9] and Yu. Petunin Ellipses
[10, 11], statistical properties for several distributions are tested.
Other popular topics about prediction sets include classification of the
methods of multivariate ordering was proposed by Barnett [12]. The concept
of the statistical depth was developed in the papers of Tukey [13], Oja [14],
Koshevoy and Mosler [15], Liu [16], Zuo and Serfling [17], Cascos [18],
Lange and Mozharovsky [19] etc.).
The methods of the construction of tolerance intervals and prediction sets
were considered by Wilks [20], Wald [21], Tukey [22], Fraser and Guttman
[23], Guttman [24], Aichison and Dunsmore [25], Bairamov and Petunin
[26], Shafer and Vovk [27], Lei et al. [28], Ndiaye and Takeuchi [29] etc.
The purpose of this section is to describe new method of construc-
tion of ellipsoidal prediction set for a cloud of the random points using
exchangeability of the points.
For more information about advanced control systems, their design and
applications we may refer to [30–34]. Modern research in theory and appli-
cations of control systems is represented there, based on contributions by
researchers from more than 16 different that report on recent developments
and new directions in advanced control systems. Also new theoretical find-
ings, industrial applications and case studies on complex engineering systems
are represented.

15.2 Petunin Ellipses and Their Statistical Properties

15.2.1 Petunin Ellipse and Ellipsoids
Let X = x1 , ..., xn be a set of random points, where xi ∈ Rd , and Ax ≤ b
is a system of linear inequalities, where A = {aij }s,d
i,j=1 . The problem is to
find the optimal ellipsoid containing at least given number of points, having
minimal volume among the similar ellipsoids and laying within the polygon
corresponding to the linear constraints Ax ≤ b.
To describe the algorithm of the construction of the Petunin ellipsoids
let us consider two cases: d = 2 and d > 2 without constraints. Essentially,
382 Petunin Ellipsoids in Automatic Control Systems Design

Figure 15.1 Petunin rectangle Π

this is the same algorithm that has geometrical features dependent from the
space dimension. Case d = 2 is more descriptive and allow simple geometric
concepts to be used for illustration.
Case d = 2. Construct a convex hull of X = {(x1 , y1 ), ..., (xn , yn )} Find
a diameter of the convex hull and its ends (xk , yk ) and (xl , yl ). Construct
a segment L connecting these ends. Find points laying at the most distance
from L: (xr , yr ) and (xq , yq ).
Construct segments L1 and L2 that are parallel to L, pass through (xr , yr )
and (xq , yq ), and have the length equals to the length of L. Construct segments
L3 and L4 that are orthogonal to L and pass through (xk , yk ) and (xl , yl ).
Segments L1 , L2 , L3 and L4 form a rectangle Π . Denote a short side by a
and a long side by b (Figure 15.1).
Make translation, rotation and shrinking with a coefficient α = ab so
  
that Π transforms to a square Π with a center (x0 , y0 ). The random points
      
are transformed to points (x1 , y1 ), (x2 , y2 ), ...(xn , yn ) ∈ Π . Find distances
        
r1 , r2 , ..., rn from (x0 , y0 ) to (x1 , y1 ), (x2 , y2 ), ...(xn , yn ) ∈ Π . Let R =
 
max(r1 , r2 , ..., rn ). Construct a circle C with the center (x0 , y0 ) and radius
     
R. Thus, (x1 , y1 ), (x2 , y2 ), ..., (xn , yn ) ∈ C (Figure 15.2).
Performing inverse transformations of this circle we obtain an ellipse E
(Figure 15.3). The computational complexity of this algorithm is equal to
computational complexity of the construction of a convex hull O(nlgn).
 15.2 Petunin Ellipses and Their Statistical Properties 383

Figure 15.2 Petunin circle C

Figure 15.3 Petunin ellipse E

Case d > 2. Construct the convex hull of X = {x1 , ..., xn }. Find

a diameter of the convex hull and its ends (xk , yk ) and (xl , yl ). Using

rotation and translation let us align the diameter along to Ox1 . Project
     
all the points (x1 , y1 ), (x2 , y2 ), ..., (xn , yn ) to the orthogonal complement

of Ox1 . Using convex hull construction, rotation and translation until the
384 Petunin Ellipsoids in Automatic Control Systems Design

orthogonal complement becomes two-dimensional we obtain a rectangle Π.

Perform the algorithm for case d = 2. Construct an axis-aligned parallelogram
 
in d dimensional space containing the images x1 , ..., xn of input points.
Transform this parallelogram to hypercube. Denote its center by x0 and find
 
distances r1 , r2 , ..., rn from it to x1 , ..., xn . Let R = max(r1 , r2 , ..., rn ).
Construct a hypersphere with the center x0 and radius R. Performing inverse
transformations we obtain an ellipsoid in the input space.
The Petunin ellipsoid have some wonderful features. First, it uniquely
arranges the points according their statistical depth because by construction
at the surface of Petunin ellipsoid only one point lays. Second, the probability
that random points from the same distribution lays within Petunin ellipse
is equal n−1
n+1 . Thus, we have not only ellipse containing a given set of n
random points with the exact probability, but also may find outliers. Peeling
these ellipsoids we may construct an influential set of points in design
space.
The main complexity problem of the construction of the Petunin ellipsoid
in a space of high dimension is the procedure of finding of two most distant
points (ends of the set diameter). As it was noted  nin [4], the complexity of
the construction of a convex hull in R is o n# 2 $ But there are some
d

effective and fast procedures [2, 5] that are not optimal but extremely smooth
the problem.

15.2.2 Petunin Ellipses and Linear Constraints

The above algorithm does not take into account the linear constraints. It
constructs an ellipse just containing all random points with given probability.
But the design of this algorithm allow easy modification so that we can select
some Petunin ellipse with the given confidence level that lays within the
polygon formed by the constraints Ax ≤ b
To do this we use a procedure proposed in [1]. Let us return to the
step in case d = 2 when we construct the concentric circles. These circles
cannot be lying within the polygon. Now, finding the side that intersects
with the outer circles (with radius ) we may construct the normal vector
to this side and shrink the circle such that it become tangent to this side.
Repeating this procedure as many times as the linear constraints are violated
we obtain a desired ellipse satisfying the linear constrains. Only drawback is
that this ellipse may have a lower confidence level since this level depends
on the points covered by the ellipse. This procedure may be extended to the
multidimensional case easily.
 15.2 Petunin Ellipses and Their Statistical Properties 385

15.2.3 Statistical Properties of Petunin‘s Ellipses

Let G be a population of random values following an unknown distribu-
tion function F and B is such that P {x ∈ B} = 1 − α, where is an
arbitrary sample element and α is a given significance level. If sample
values x1 , x2 , ..., xn are exchangeable identically distributed random values
following absolutely continuous distribution function, then [9],
) ) ** j−1
p xn+1 ∈ x(i) , x(j) =
n+1
where xn+1 is the next sample value from G, and xi , xj are ith and jth order
statistics.
The following theorems hold [8].
Theorem 15.1. If x1 , x2 , ..., xn+1 are exchangeable random values
6 following
7
absolutely continuous joint distribution function such that P xk = x(j) =
0 if k ∓ m, then
1
P {xk ≥ x1 , ..., xk ≥ xk−1 , xk ≥ xk+1 , ..., xk ≥ xn+1 } = .
n+1

Theorem 15.2. If x1 , x2 , ..., xn+1 are exchangeable random values

6 following
7
absolutely continuous joint distribution function such that P xk = x(j) =
0 if k ∓ m, and x(1) ≤ x(2) ≤ ... ≤ x(n) the a variational series, then
) ) **
P xn+1 ∈ x(i) , x(j) = n+1 1
.
) ) ** j−1
Corollary. p xn+1 ∈ x(i) , x(j) = n+1 ∀0 ≤ i < j ≤ n + 1, where
η0 = −∞ηn+1 = +∞.
Theorem 15.3. If x1 , x2 , ..., xn are exchangeable random vectors from G,
En is the Petunin ellipsoid, containing x1 , x2 , ..., xn , and xn+1 ∈ G, then
p(xn+1 ∈ E) = n+1
n
.
Corollary 15.1. The significance level of the Petunin ellipsoid is less than
0,05, if n > 19.
Corollary 15.2. If n>19, then the area of the Petunin ellipsoid can be
decreased without increasing of the significance level, deleting farthest points
r(l) , ·, r(n) , if n >19.
This algorithm has remarkable features: 1) only one point always lays at the
surface of the Petunin ellipsoid; 2) the Petunin ellipsoid contains n point with
probability n−1
n+1 ; 3) in practice, to guarantee the significance level which is
less than 0.05 a set of points may contain only 20 random samples.
386 Petunin Ellipsoids in Automatic Control Systems Design

15.2.4 Numerical Experiments

Hypothetically, the more points we have the more confidence level of the
Petunin ellipsoid covering all the point, and therefore, higher confidence level
of the Petunin ellipsoid satisfying the linear constraints. In order to test this
hypothesis, we have carried out several numerical experiments with two-
dimensional samples following the uniform and Gaussian distributions and
satisfying constraints that are defined by a square, pentagon and hexagon.
We suppose that the more circular-like a polygon defining constraint the less
points we loss during peeling the Petunin ellipsoid. We 100 times generated
N two-dimensional samples, where N = 40, 50, ..., 100, constructed the
optimal Petunin ellipses covering part of these points and satisfying a system
of linear constraints and averaged the number of points that are not covered
by the Petunin ellipse satisfying a system of linear constraints (Table 15.1–
15.2). The purpose of these experiments where to determine the degree
of decreasing of the confidence level of ellipses depending on the kind if
constraints (Figures 15.1–15.7).

Table 15.1 The number of uniformly distributed points dropped out and the significance
level of the Petunin ellipse satisfying square constraints
Initial Number of Dropped Initial Significance Result Significance
Number out Points Level Level
40 19 0.976 0.955
50 7 0.980 0.977
60 9 0.984 0.981
70 19 0.986 0.981
80 24 0.988 0.982
90 50 0.989 0.976
100 66 0.990 0.971

Table 15.2 The number of Gaussian distributed points dropped out and the significance level
of the Petunin ellipse satisfying square constraints
Initial Number of Dropped Initial Significance Result Significance
Number out Points Level Level
40 7 0.976 0.971
50 14 0.980 0.973
60 12 0.984 0.983
70 13 0.986 0.985
80 8 0.988 0.986
90 8 0.989 0.988
100 9 0.990 0.989
 15.2 Petunin Ellipses and Their Statistical Properties 387

Figure 15.4 Uniformly distributed points with square constraints

Figure 15.5 Gaussian distributed points with square constraints

Figure 15.6 Uniformly distributed point with pentagonal constraints

As we see in the uniform case, when the sample size is greater than 40
the decreasing of the confidence level of the Petunin ellipses after peeling is
varied in the region (0.03; 0.13). When the sample size is large (90 or 100) we
388 Petunin Ellipsoids in Automatic Control Systems Design

Figure 15.7 Uniformly distributed point with hexagonal constraints

have clear gift in the number of points (55%–66%) without significant loss of
the confidence level of the ellipse.
As we see in the Gaussian case, when the sample size is greater than 50
the decreasing of the confidence level of the Petunin ellipses after peeling is
varied in the region (0.01; 0.02). However, we had not achieved a significant
decreasing of the number of representative points. Thus, we may suppose that
this effect is connected with the kind of distribution.
Figure 15.8 illustrates changing of area of Petunin ellipses after drop-
ping one point per step. As far as the Petunin ellipses covering points are
concentric the area is monotonically decreasing. Absence of abrupt jumps
in these curves justify that the sample is homogeneous and fill the region
densely.

Figure 15.8 Area of Petunin ellipses covering 100 uniformly and Gaussian distributed points
after dropping one point per step
 15.2 Petunin Ellipses and Their Statistical Properties 389

15.2.5 Prediction Sets

Generalization of training sample and forecast future values is the primary
problem of machine learning. The main tools for solving this problem are
“predictive sets” that are also named “tolerance regions” and “minimum
volume sets”. Consider the general problem of constructing conformal pre-
diction set: given i.i.d. points x1 , x2 , ..., xn ∈ Rd find a prediction set
E(x1, x2 , ..., xn ) ⊂ Rd such that p(xn+1 ∈ E) ≥ 1 − α, where 0 < α < 1
is a given significance level, thus p(xn+1 ∈ E) is the confidence level of the
predictive set.

15.2.6 Hill’s Assumption A(n)

Let x1 , x2 , ..., xn be a sample drawn by the simple sampling from a pop-
ulation following absolutely continuous distribution F. Arrange it in the
increasing order and construct the variance series x(1) ≤, x(2) ≤ ...xn ,
where x(i) is ith order statistics. The order statistics x(1) , x(2) , ..., x(n) in
contrast to initial sample values are dependent and have different distribu-
tions. The distribution Fk (u) of the kth order statistics x(k) is Fk (u) =
n
Cni [F (u)]i [1 − F (u)]n−i , where πeF (u) = p(x ≤ u). The Hill’s
i=k
assumptions An [9] states that if xn+1 drawn from the population following
distribution then
) ) ** j−1
p xn+1 ∈ x(i) , x(j) = , j > i.
n+1
The Hill’s assumption was proved in papers of Yu.I. Petunin et al. [10, 11]
for i.i.d. and exchangeable random variables, respectively.
Let us consider the proof provided in [21]. If random variables and are
independent, then
∞
p (ξ < η) = Fξ (u)dFη (u),
−∞

where Fξ (u) and Fη (u) are the distribution functions of ξ and η, respectively.
The probability density of ith order statistics is:
!
n !
n
fk (u) = Cni [F (u)]i [1 − F (u)] n−i
= Gi (u).
i=k i=k
390 Petunin Ellipsoids in Automatic Control Systems Design

Hence,

!
n

Fk (u) = Gi (u)
i=k

Gk (u) = Cnk [k(F (u))k−1 (1 − F (u))n−k f (u)−
− (F (u)) (n − k)(1 − F (u))n−k−1 f (u)].
k


' (
Gk+1 (u) = Cnk+1 (F (u))k+1 (1 − F (u))n−k−1
= Cnk+1 [(k + 1)(F (u))k (1 − F (u))n−k−1 f (u)−
− (F (u))k+1 (1 − F (u))n−k−2 f (u)].

The second term annihilates with the first term:

n!(n − k) n!(k + 1)
−Cnk (n − k) + Cnk1 (k + 1) = +
(n − k − 1)!k! (n − k − 1)!(k + 1)!
n! n!
=− + = 0.
(n − k − 1)!k! (n − k − 1)!k!
The last term of previous sum is zero
(F (u))n (1 − F (u))0 (n − n)f (u) = 0.
Thus,
k−1
fk (u) = nCn−1 [F (u)]k−1 [1 − F (u)]n−k f (u).
Let us find p(x < x(i) ) and p(x < x(j) ). Using the above equations, we
have
∞ ∞
(i)
p(x < x ) = n+1
F (u)dFi (u) = nCn−1 [F (u)]i [1 − F (u)]n−k dF (u)
−∞ −∞
1
i−1
= nCn−1 v i (1 − v)n−1 dv
0
.
It is well-known that,
1
Γ(p)Γ(q) (p − 1)!(q − 1)!
xp−1 (1 − x)q−1 dx = =
Γ(p + q − 1) (p + q − 1)!
0
 15.2 Petunin Ellipses and Their Statistical Properties 391

Thus,
1
xi+1−1 (1 − x)n−i+1−1 dx = B (i + 1, n − i + 1)
0
Γ(i + 1)Γ(n − i + 1) Γ(i + 1)Γ(n − i + 1) i!(n − i)!
= = =
Γ(i + 1 + n + 1 − i) Γ(n + 2) (n + 1)!
j!(n − i)!
p (xn+1 < xi ) = ni−1
n−1 =
(n + 1)!
(n − 1)!j! (n − j)!
=n =
(n − 1 − j + 1)! (j − 1)! (n + 1)!
n (n − 1)!(j − 1)j! (n − j)! j
= =
(n − j)!(j − 1)!n (n + 1) (n − 1)! n+1
Therefore,
) ) ** j i j−i
p xn+1 ∈ x(i) , x(j) = p (xn+1 < xi ) = − =
n+1 n+1 n+1
Thus, if a random variable x does not depend on x1 , x2 , ..., xn and
it is drawn by sample sampling from the same population following the
distribution F (u), then
) ) ** n − 1
p x ∈ x(1) , x(n) =
n+1
) *
Remark 15.1. The confidence level of the tolerance interval J = x(1) , x(n)
n+1 , thus for n ≥ 39 the confidence level of this interval is less than 0,05.
is n−1

15.2.7 Testing Statistical Properties for Petunin Ellipses

To test the proposed algorithm we used two-dimensional samples following
the uniform distribution in [0,100]× [0,100] and Gaussian distribution N(μ, ),
where μ = (50,50) and = diag(10, 10). We computed the Petunin ellipses
covered 40 points distributed in and constructed the curve of area of every
concentric ellipse.
At Figure 15.9 we see that six (35–40) outer points are outliers because
the area of the corresponding ellipses significantly decreases. Also, we see
natural clustering of the almost equiprobable points. The borders of the
clusters are the points that are not outliers but the corresponding ellipse area
392 Petunin Ellipsoids in Automatic Control Systems Design

Figure 15.9 Ellipse covering point with uniform (left) and Gaussian (right) distribution

Figure 15.10 Areas of the ellipses covering point with uniform (left) and Gaussian (right)
distribution

significantly decreases. We detected the following clusters of points: 1–5,

6–8, 9–16, 17–23, 24–28, 29–34, 35–38, and 39–40.
At Figure 15.10 we see that three (38–40) outer points are outliers because
the area of the corresponding ellipses abruptly decreases. More flat curve
comparing the curve for the uniform distribution is a consequence of more
compact arranging of points with the Gaussian distribution. The number of
clusters of equiprobable points following the Gaussian distribution is less
than for the uniformly distributed points. We detected the following clusters
of points: 1–4, 5–14, 15–19, 17–32, and 33–37.
 15.2 Petunin Ellipses and Their Statistical Properties 393

The Petunin ellipses accurately reflect the statistical properties of distribu-

tions. More dispersed distribution (uniform) demonstrate the greater number
of clusters of equiprobable points and greater area of ellipses, but more
compact distribution (Gaussian) produce the less number of clusters and less
area of ellipses.
Also, we tested whether the Petunin ellipses satisfy the condition on the
coverage probability P (xn+1 ∈ E) = n−1 n+1 . We generated 40 points from the
uniform distribution in [0,100]× [0,100] and constructed the corresponding
covering Petunin ellipse E. Then, we generated next 100 points from the
same distribution and found the relative frequency of the event P (xn+1 ∈ E)
We repeated these trials 20 times and averaged the relative frequencies.
To estimate how the average relative frequency is close to the expected
probability n−1
n+1 we used the Wilson confidence interval and the 2s–rule.
The results are provided in Table 15.1. The expected probability is covered
both by the Wilson interval and by the 2s–interval, therefore, the experiment
confirms the properties of the Petunin ellipse.

Table 15.3 The results of testing confidence level of the Petunin ellipse for uniform
distribution
Observable Relative
Trial Frequency Descriptive Statistics
1 0.98 Expected probability 0.9048
2 0.95 Mean 0.9616
3 0.96 Standard Error 0.0073
4 0.92 Median 0.9700
5 0.99 Mode 0.9800
6 0.99 Standard Deviation 0.0318
7 0.94 Sample Variance 0.0010
8 0.95 Kurtosis 0.7694
9 0.98 Skewness –0.9680
10 1 Range 0.1200
11 1 Maximum 1.0000
12 0.98 Minimum 0.8800
13 0.97 Geometric Mean 0.9610
14 0.88 Harmonic Mean 0.9605
15 0.95
16 0.98
17 0.99
18 0.98
19 0.94
(Continued)
394 Petunin Ellipsoids in Automatic Control Systems Design

Table 15.3 Continued

Observable Relative
Trial Frequency Descriptive Statistics
20 0.92
Expected probability 0.9048
Wilson interval 0.7248 0.8880 1.0000
2s-interval 0.8988 0.9625 1.0000

15.3 Conclusion
The proposed algorithm of constructing optimal Petunin ellipsoid in design
space is effective and statistically robust. It allows decreasing of the point
in design space without loss of confidence level of the ellipsoids. This algo-
rithm demonstrates robustness to outliers, high confidence level with modest
number of covering points and useful possibility to arrange the point on the
probability to cover them by the ellipsoid.
The proposed algorithm of constructing conformal prediction sets using
Petunin ellipsoid is effective and has high performance. It has theoretically
precise confidence level and allows unambiguous multivariate points arrang-
ing, computing statistical depth every points, finding median of the set, and
detecting outliers. The experiments confirmed the theoretical properties of
the Petunin ellipses.
The Petunin ellipsoid is useful tool for arrange multivariate data and
finding anomalies using statistical depth. Thus, the future research directions
within the domain of the topic are developing computationally effective
algorithms for construction of high-dimensional conformal prediction sets
using genetic and other algorithms and inventing new multivariate statistical
tests for homogeneity of multivariate samples in high-dimensional spaces.

References
[1] A. Bedrintsev, V. Chepyzhov, “Description of the design space by
extremal ellipsoids in data representation problems”, J. Commun. Tech-
nol. El., vol. 61, pp. 688–694, 2016.
[2] S. Har-Peled, “A practical approach for computing the diameter of a
point set”, In: Symposium on Computational Geometry (SOCG’2001),
pp. 177–186, 2001.
[3] S. Lyashko, D. Klyushin, V. Alexeenko, “Multivariate ranking using
elliptical peeling”, Cybern. Syst. Anal., 49, pp. 511–516, 2013.
 References 395

[4] S. Lyashko., D. Klyushin., V. Semenov, M. Prysiazhna, M. Shlykov,

“Nonparametric Ellipsoidal Approximation of Compact Sets of Ran-
dom Points”, Chapter in: Optimization and Applications in Control
and Data Sciences (ed. B.Goldengorin), Springer, Optimization and Its
Applications, vol. 115, pp. 327–340, 2016.
[5] G. Malandain, J.-D. Boissonnat, “Computing the Diameter of a Point
Set. International Journal of Computational Geometry and Applica-
tions”, IJCGA, vol. 12, pp. 197–208, 2002.
[6] Yu. Petunin, B. Rublev, “Pattern recognition with the help of
quadratic discriminant function”, J. Math. Sci., vol. 97, pp. 3959–3967,
1999.
[7] S.I. Lyashko, B.V. Rublev, “Minimal Ellipsoids and Maximal Simplexes
in 3D Euclidean Space”, Cybernetics and Systems Analysis, vol. 39,
pp. 831–834, 2003.
[8] S.I. Lyashko, D.A. Klyushin, V. V. Alexeyenko, “Multivariate ranking
using elliptical peeling”, Cybernetics and Systems Analysis, vol. 49,
pp. 511–516, 2013.
[9] B.M. Hill, “Posterior distribution of percentiles: Bayes” theorem for
sampling from a population”, J. Am. Stat. Assoc., vol. 63, pp. 677–691,
1968.
[10] I. Madreimov, Yu.I. Petunin, “Characterization of the uniform distri-
bution using order statistics”, Teor. Veroyatn. Mat. Statist., vol. 27,
pp. 96–102, 1982.
[11] R.I. Andrushkiw, D.A. Klyushin, Yu. I. Petunin, V. N. Lysyuk. “Con-
struction of the bulk of general population in the case of exchangeable
sample values”, Proc. Intern. Conf. of Mathemat. and Engineer. Techniq.
in Medicine and Biolog. Sci. (METMBS’03), Las Vegas, Nevada, USA,
pp. 486–489, 2003.
[12] V. Barnett, “The ordering of multivariate data”, Journal of the Royal
Statistical Society. Series A (General), vol. 139, pp. 318–355, 1976.
[13] J.W. Tukey, “Mathematics and the picturing of data”, Proc. of the Inter-
national Congress of Mathematician, Montreal, Canada, pp. 523–531,
1975.
[14] H. Oja, “Descriptive statistics for multivariate distributions”, Statistics
and Probability Letters, vol. 1, pp. 327–332, 1983.
[15] G. Koshevoy, K. Mosler, K. “Zonoid trimming for multivariate distribu-
tions”, Annals of Statistics, vol. 25, pp. 1998–2017, 1997.
[16] R.J. Liu, “On a notion of data depth based on random simplices”, Annals
of Statistics, vol. 18, pp. 405–414, 1990.
396 Petunin Ellipsoids in Automatic Control Systems Design

[17] Y. Zuo, R. Serfling, “General notions of statistical depth function”,

Annals of Statistics, vol. 28, pp. 461–482, 2000.
[18] I. Cascos, “Depth function as based of a number of observation of a
random vector”, Working Paper 07–29, Statistic and Econometric Series
07, vol. 2, pp. 1 28, 2007.
[19] T.I. Lange, P.F. Mozharovsky, “Determination of the depth for mul-
tivariate data sets”, Inductive simulation of complex systems, vol. 2,
pp. 101–119, 2010.
[20] S.S. Wilks, “Statistical Prediction With Special Reference to the Prob-
lem of Tolerance Limits”, Annals of Mathematical Statistics, vol. 13,
pp. 400–409, 1942.
[21] A. Wald, “An extension of Wilks method for setting tolerance limits”,
The Annals of Mathematical Statistics, vol. 14, pp. 45–55, 1943.
[22] J. Tukey, “Nonparametric estimation. II. Statistical equivalent blocks
and multivarate tolerance regions”, The Annals of Mathematical Statis-
tics, vol. 18, pp. 529–539, 1947.
[23] D.A.S. Fraser, I. Guttman, “Tolerance regions”, The Annals of Mathe-
matical Statistics, vol. 27, pp. 162–179, 1956.
[24] I. Guttman, “Statistical Tolerance Regions: Classical and Bayesian Grif-
fin”, Hartigan, J., editor. London.; Clustering Algorithms John Wiley;
New York: 1970.
[25] J. Aichison, I.R. Dunsmore, “Statistical Prediction Analysis”
Cambridge University Press, 1975.
[26] I.G. Bairamov, Yu. I. “Invariant confidence intervals for the main mass
of values from the distribution in a population”, J Math Sci, vol. 66,
pp. 2534–2538, 1993.
[27] G. Shafer, V.A. Vovk, “A Tutorial on Conformal Prediction”, Journal of
Machine Learning Research, vol. 9, pp. 371–421, 2008.
[28] J. Lei, J. Robins, L. Wasserman, “Distribution Free Prediction Sets”,
Journal of the American Statistical Association, vol. 108, pp. 278–287,
2013.
[29] E. Ndiaye, I. Takeuchi, “Computing Full Conformal Prediction Set
with Approximate Homotopy”, NeurIPS Proceedings, pp. 1384–1393,
2019.
[30] V.M. Kuntsevich et al. (Eds). Control Systems: Theory and Applica-
tions. Series in Automation, Control and Robotics, River Publishers,
Gistrup, Delft, 2018.
 References 397

[31] Y.P. Kondratenko, V.M. Kuntsevich, A.A. Chikrii, V.F. Gubarev

(Eds). Advanced Control Systems: Theory and Applications. Series in
Automation, Control and Robotics, River Publishers, Gistrup, 2021.
[32] Y.P. Kondratenko, A.A. Chikrii, V.F. Gubarev, J. Kacprzyk (Eds).
Advanced Control Techniques in Complex Engineering Systems: The-
ory and Applications. Dedicated to Professor Vsevolod M. Kuntsevich.
Studies in Systems, Decision and Control, Vol. 203. Cham: Springer
Nature Switzerland AG, 2019.
[33] V.M. Kuntsevich, “Estimation of Impact of Bounded Perturbations on
Nonlinear Descrete Systems”, In: V.M. Kuntsevich, et al. (Eds). Control
Systems: Theory and Applications. Series in Automation, Control and
Robotics, River Publishers, Gistrup, Delft, 2018, pp. 3–15.
[34] J. Kacprzyk, et al. A Status Quo Biased Multistage Decision Model
for Regional Agricultural Socioeconomic Planning Under Fuzzy Infor-
mation. In: Y.P. Kondratenko, A.A. Chikrii, V.F. Gubarev, J. Kacprzyk,
(Eds) Advanced Control Techniques in Complex Engineering Systems:
Theory and Applications. Dedicated to Prof. V.M.Kuntsevich. Studies
in Systems, Decision and Control, Vol. 203. Cham: Springer Nature
Switzerland AG, 2019, pp. 201–226.
 16
On Real-Time Calculation of the Rejected
Takeoff Distance

V.I. Vyshenskyy, A.A. Belousov, and V.V. Kuleshyn

V.M. Glushkov Institute of Cybernetics of National Academy of Sciences

of Ukraine, Kyiv, Ukraine
E-mail: [email protected]; [email protected];
[email protected]

Abstract
A rejected takeoff is the situation when it is decided to cancel the takeoff
of an airplane. When a crucial obstacle to the takeoff occurs (e.g. engine
failure), the crew compares the current velocity with the “decision-making
velocity”. Based on this velocity one can estimate whether the aircraft can
safely stop within the available distance of the interrupted flight. These values
are taken from the Aircraft Flight Manual before the takeoff. But the real
rejected takeoff distances can differ from the pre-calculated values. The goal
of our investigation is to obtain real-time estimations of the distance to full
stop. The proposed real-time algorithm of the calculation of the distance of
the interrupted flight uses the step-wise interpolation of the thrust. Using
this interpolation, the Riccati equation of the aircraft motion has explicit
solutions. Besides using known time intervals of interpolation, the time to
reach a given speed can be explicitly calculated. To check and study the
proposed algorithm for estimating the rejected takeoff distance in real time,
a computer simulation system was created. To verify the adequacy of the
algorithm, the emulation results were compared with the values obtained
from the real aircraft nomograms. The algorithm takes into account different
aircraft and environment parameters (thrust, velocity, temperature etc.) before
the moment of decision-making. As a result, the decision about the rejected
takeoff will be more reliable and safe.

399
400 On Real-Time Calculation of the Rejected Takeoff Distance

Keywords: rejected takeoff, braking distance, real-time algorithm, the

Riccati equation.

16.1 Introduction
A rejected takeoff is the situation when it is decided to cancel the takeoff of an
airplane. Analysis of accidents and serious incidents in civil aviation shows
that an aircraft exiting the runway and terminal safety strip does not happen
often when the takeoff is cancelled, but it is extremely dangerous. The reasons
for these accidents and incidents may be the following: too late actions of
the crew to stop takeoff; inconsistency of the conditions for the interrupted
takeoff with the calculated ones. In the first case, the pilot or crew reacts to the
appearance of hazard factors with a time delay greater than the allowable one.
In the second, the conditions for a rejected takeoff do not correspond to those
that were used in the preliminary calculations. In both cases, the aircraft does
not have enough time to stop within the available distance of an interrupted
takeoff (ASDA, Accelerate-Stop Distance Available), since the real distance
of an interrupted takeoff (ASD, Accelerate-Stop Distance) exceeds ASDA.
When a crucial obstacle to the takeoff occurs (e.g. engine failure), the
crew has just several seconds to analyse the situation and to make a “go–
no go” decision–to continue a takeoff, or to abort it. During this time, the
aircraft speed increases from vEF (engine failure velocity) to vEF + Δv. The
crew compares current velocity to the “decision-making velocity” V1 . Based
on V1 velocity, it is possible to estimate if the airplane still can safely stop
within the available distance of the airport. If the current speed is less than
the “decision-making velocity”, the crew makes a decision to interrupt the
takeoff; otherwise, it is necessary to continue the takeoff, avoiding overrun. It
is assumed that if the crew decides to terminate the takeoff at a speed less than
or equal to V1 , the aircraft will stop within the ASDA. The velocity V1 and
ASDA distance are calculated in advance using the “Aircraft Flight Manual”
[1]. However, the real rejected takeoff distances (the length of the braking
section until the aircraft stops completely) can differ from the pre-calculated
values used in flight manuals and nomograms due to different hazard factors
[2, 3]. Among these factors, we should consider possible differences between
the actual characteristics of the engines, the take-off weight of the aircraft and
the conditions of the airfield (especially friction during takeoff and braking)
from the characteristics used in calculating the airplane flight curves and
nomograms. Therefore, a more precise real-time calculation of the rejected
takeoff dista2nce performed by an automatic control system of the aircraft
can help the crew to make accurate and safer “go–no go” decision.
 16.2 Dynamic Model of Aircraft Movement along the Runway 401

This chapter continues some researches of previous fundamental and

applied works [4–13]. These publications were devoted to the researches
in complex control systems, conflict-controlled processes, dynamic games
of approach, aircraft control during takeoff. The goal of our investigation
is to obtain real-time estimations of the distance to full stop. The proposed
algorithms use step interpolation of thrust, and also of other aircraft parame-
ters. With such interpolation, the Riccati equation of the plane’s motion has
explicit solutions. In addition to using interpolation time intervals, the time
to reach a given speed can be explicitly calculated, because some parameters
change abruptly upon reaching a certain speed. Mathematical modelling of
the rejected takeoff distance calculation makes it possible to evaluate the
accuracy and speed of the algorithm, as well as its relevance to the real data
of nomograms and airplane flight manual. Simulations can also investigate
the influence of various hazard factors on rejected takeoff parameters.

16.2 Dynamic Model of Aircraft Movement along the

Runway
The aircraft motion on a runway can be described by differential equa-
tions [14]
dv
m = P cos αP − Q − mg sin θ + f (Y − mg) cos θ (16.1)
dt
dl
= v, (16.2)
dt
v(0) = 0, l(0) = 0, (16.3)

where t–time, v–velocity, l–distance on the airport runway, m–mass, P –thrust

2
of all aircraft engines, αP –aircraft engine angle of attack, Q = cx ρv2 S–
2
aircraft drag force, Y = cy ρv2 S–aircraft lift, cx –aircraft drag coefficient, cy –
aircraft lift coefficient, S–relevant surface area, ρ–air density, g–acceleration
of gravity, θ–runway angle, f –friction coefficient.
Equation (16.1) is a Riccati equation
dv
= −Λv 2 + G, (16.4)
dt
ρS
Λ= · (cx − f · cy ) , (16.5)
2m
cos αP
G= P − g(sin θ − f cos θ), (16.6)
m
402 On Real-Time Calculation of the Rejected Takeoff Distance

where in (16.4) both Λ and G depend on aircraft and environment parameters

(aircraft mass, friction coefficient, air density, lift and drag coefficients, etc.),
but only G depends on the thrust.

16.3 Analytical Solution of the Aircraft Motion Equation

Consider the functionx(t) = eΛl(t) . For this function, we have

deΛl dl
= ΛeΛl = ΛeΛl v,
dt dt
 Λl  '
d2 eΛl de Λl dv
) *(
= Λ· v + e = Λ· Λe Λl 2
v + e Λl
−Λv 2
+ G = ΛGeΛl .
dt2 dt dt
√
1. IfΛG < 0, thenẍ = −ω 2 x, whereω = −ΛG. This is the classical
equation of small oscillations of a pendulum [15]. General solution of
this equation is
x(t) = C1 sin ω t + C2 cos ω t, (16.7)
with constants C1 , C2 that can be obtained
√ from the initial conditions.
2. If ΛG > 0, then ẍ = γ 2 x, where γ = ΛG. This is the equation of the
“inverted pendulum” [15] with general solution

x(t) = C1 sh γ t + C2 chγt, (16.8)

−t −t
sh t = e −e
t t
2 , ch t = e +e
2 - hyperbolic sine and cosine.
3. If ΛG = 0, we obtain a trivial equation with general solution that
describes the motion by inertia

x(t) = C1 t + C2 . (16.9)

16.4 Engine Thrust Modelling During Switching

off and Reverse
Let us analyse the actions of pilots and the change in thrust of the engines
during rejected takeoff. Before the cause of the aborted takeoff (for example,
engine failure) is detected, the airplane travels the distance l0 and accelerates
to the speed vEF . Then the pilot spends some time δ identifying the cause
and making a decision. If the current airplane speed is less than the decision-
making velocity V1 , the process of the rejected takeoff usually starts. The
thrust of the engines at the time t0 +δ is practically equal to the thrust P0 at the
 16.4 Engine Thrust Modelling During Switching off and Reverse 403

time t0 . At this time, the engine control levers are moved to the “Small Gas”
position. As a result, the thrust of the engines is reduced to almost zero in 1.5–
2 seconds. Then the breaking systems are activated, including reversal thrust.
When the reverse is turned on, the negative thrust of the engines changes
from 0 to the full reverse thrust U . This process requires about 7–11 seconds.
It is assumed that the transient processes end when they reach 0.95 of the
steady-state values [16].
Complex mechanical-mathematical and thermodynamic relations describe
the transients of modern turbojet engines. However, from the point of view
of control theory, the integral effect of engine thrust on an aircraft during
the transient process is described with good accuracy by linear differential
equations [16, 17].
The equation for decreasing the takeoff thrust of the engine to zero can be
written as follows [17]
dP P
=− ,
dt T1
−t/
P (t) = P0 · e T1 , (16.10)
where the value T1 is called the engine shutdown time constant and is an
important characteristic of a particular type of turbojet engine. Function
(16.10) satisfies the relation

P (3T1 ) = P0 · e−3 ≈ 0, 05P0 , (16.11)

  −t/
3T1 3T1 ) *
P (t) dt = P0 e T1 dt = P0 · T1 1 − e−3 ≈ 0, 95·
0 0
 T1
P0 · T1 = 0, 95 · P0 dt.
0

We can substitute the thrust function (16.10) in the Riccati Equation (16.4)
and numerically find the distance travelled by the aircraft during the transition
process. However, given the relation (16.11), instead of the exponential thrust
function (16.10) it is possible to use the step-wise constant function

P0 , 0 ≤ t < T1 ,
P (t) =
0, T1 ≤ t ≤ 3T1 ,

that fairly well approximates the impact of thrust (16.10).

404 On Real-Time Calculation of the Rejected Takeoff Distance

Figure 16.1 Dependence of engine thrust on time

Similarly, we can consider the process of the engine reversal from zero to
the nominal reverse thrust, which is described by the function
t/ −3
P (t) = U · e T2 , 0 ≤ t ≤ 3T2 , (16.12)

where T2 is the reverse time constant.

Thus, the processes (16.12) of switching off the thrust of the engine
and the activating of reverse mode are proposed to be emulated by a step
function
⎧
⎨P0 , t0 ≤ t < t1 = t0 + δ + T1 ;
⎪
P (t) = 0, t1 ≤ t < t2 = t1 + 2T1 + 2T2 ; (16.13)
⎪
⎩
U, t2 ≤ t.

Dependence of engine thrust on time described by (16.10) and (16.12), as

well as their emulated values described by (16.13), are shown on Figure 16.1.
Note that the actual shutdown of the takeoff thrust starts at the time t0 +δ,
and the start of reverse mode at the time t0 + δ + 3T1 .

16.5 Rejected Takeoff Distance Calculation

Denote li = l(ti ), vi = v(ti ), Δi = ti − ti−1 , Δli = li − li−1 , i = 1, 2, 3.
Determine the distance travelled by the aircraft during time interval
[t0 , t1 ] with constant thrust P0 . In this case, the value G > 0, and the general
solution has the form (16.8)
√
eΛl(t) = C1 shγt + C2 chγt, γ = ΛG,
 16.5 Rejected Takeoff Distance Calculation 405

its derivative
deΛl(t)
= ΛeΛl(t) v(t) = C1 γchγt + C2 γshγt.
dt
We can find the constants C1 and C2 from the initial conditions l(0) = 0
and v(0) = v0 :
eΛ·0 = C1 sh(0) + C2 ch(0) = C2 ,
ΛeΛ·0 v0 = C1 γch(0) + C2 γsh(0) = C1 γ,
so we have
Λv0
C1 = , C2 = 1.
γ
We get the relation between the distance Δl1 and the speed v1 over time
Δ1 = δ + T 1
Λv0
eΛΔl1 = sh(γΔ1 ) + ch(γΔ1 ),
γ
ΛeΛΔl1 v1 = Λv0 ch(γΔ1 ) + γsh(γΔ1 ),
thus  
1 Λv0
Δl1 = ln sh(γΔ1 ) + ch(γΔ1 ) , (16.14)
Λ γ
' γ (
v1 = e−ΛΔl1 v0 ch(γΔ1 ) + sh(γΔ1 ) . (16.15)
Λ
During time interval [t1 , t2 ] (with zero thrust) and [t2 , t3 ] (with reverse
thrust) we have Λ > 0, G < 0, and general solution is (16.7)

eΛl(t) = C1 sin ω t + C2 cos ω t, (16.16)

√
where ω = −ΛG. Its derivative

ΛeΛl(t) v(t) = ω C1 cos ω t − ω C2 sin ω t. (16.17)

The constants C1 , C2 can be found from initial conditions l(0) = 0 and

v(0) = v2 , that is eΛ·0 = C2 , ΛeΛ·0 v2 = ω C1 , so
Λv2
C1 = , C2 = 1.
ω
Then, substituting the condition of stop v = 0 into (16.17), we obtain
Λv2
tgω Δ3 = ,
ω
406 On Real-Time Calculation of the Rejected Takeoff Distance

and the time to full stop

 
1 Λv2
Δ3 = arctg .
ω ω
We can find the distance Δl3 from (16.16)
Λv2 1 <
eΛΔl3 = sin ω Δ3 + cos ω Δ3 = = 1 + tg 2 (ω Δ3 )
ω cos ω Δ3
> >
2
Λ v 2 2
Λv2
= 1 + 22 = 1 −
ω G
so  
1 Λv22
Δl3 = ln 1 − . (16.18)
2Λ G
Similarly, we can find the distance Δl2 during time interval [t1 , t2 ].
Thus, taking into account all three stages of braking, the total distance
travelled by the aircraft during an emergency stop will be l0 +Δl1 +Δl2 +Δl3 .
If this distance is within the length of the runway, then we assume that a safe
interrupted takeoff is possible. The total time spent on the interrupted takeoff
will be t0 + δ + 3T1 + 2T2 + Δ3 .

16.6 Interpolation of Aircraft Parameters using Constant

Values
Section 16.5 proposes an algorithm for estimating the distance of interrupted
takeoff in real time, which uses stepwise thrust interpolation. However, the
real process of acceleration and interrupted takeoff is much more compli-
cated. Engine thrust during acceleration is not constant. For example, An–124
aircraft at the beginning of the takeoff uses only 0.7 of the nominal thrust
during first 3 seconds. So we assume that P (t) = 0.7P0 , 0 < t < t∗ ,
P (t∗ ) = P0 . Engine thrust depends on speed, P = P (v) – it decreases when
the speed increases. After deciding to stop the takeoff, in addition to switch-
ing to reverse thrust, the crew increases the aerodynamic drag and reduces the
lift of the aircraft, using interceptors. As a result, aerodynamic coefficients
cx , cy in (16.1) are changed. Wheel brakes are also used and the friction
coefficient f changes. The reverse thrust decreases gradually with decreasing
speed, and when the speed of the aircraft reaches a certain value, the reverse
thrust decreases abruptly to the value of the so-called small gas reverse. Thus,
when emulating the motion of the aircraft using time intervals with constant
16.7 Estimation of Speed, Path and Time using Fixed Parameters of the Aircraft 407

values of the parameters, we can assume that some values change at fixed
known moments of time, and at these time intervals Δi = ti+1 −ti we need to
estimate li+1 = l(ti+1 ), vi+1 = v(ti+1 ) using known ti , ti+1 , l(ti ), v(ti ).
We assume that engine failure occurs at some speedvEF , and after that, the
rejecting takeoff begins, with known time parameters δ, T1 , T2 etc. We
do not know the exact time tEF of the engine failure, this time should be
calculated, but we know that P (tEF ) = P (vEF ). As a result, during the
time interval [t∗ , tEF ] we can interpolate the thrust with constant value
1 ∗
2 [P0 + P (vEF )]. For more accuracy, we can split [t , tEF ] into two or more
parts, e.g. [t∗ , τ ], [τ, tEF ], where v(τ ) = 2 [v(t∗ ) + vEF ], and we can use
1

two constant values 12 [P0 + P (v(τ ))] and 12 [P (v(τ ) + P (vEF )]. Similarly,
we assume that we know the time when the coefficients cx , cy change their
values – it takes 1–2 second after start of the reverse thrust. The time of small
gas reverse is not known, and it must be calculated during emulation.

16.7 Estimation of Speed, Path and Time using Fixed

Parameters of the Aircraft
Equation (16.4) with constant Λ, G and initial conditions
v(t0 ) = v0 , l(0) = 0, (16.19)
allows to obtain both explicit dependences of path and velocity on time,
l(t), v(t), and also analytically define the dependency of time on velocity
t(v). We will use explicit solutions (16.7) and (16.8). The case ΛG = 0 does
not occur in practice during aircraft movement.
Consider the function x(t) = eΛl(t) . For this function we have
dx dl
= ΛeΛl = Λxv, (16.20)
dt dt
 Λl  '
d2 x de Λl dv
) *(
= Λ · v + e = Λ · Λe Λl 2
v + e Λl
−Λv 2
+ G = ΛGx.
dt2 dt dt
√
Let ΛG > 0, and denote γ = ΛG. Then
x(t) = C1 shγt + C2 chγt. (16.21)
Taking into account the initial conditions, we can obtain explicit depen-
dence of time on speed [10]
 
1 1+z Λγ(v0 − v)
t= ln , z= 2 . (16.22)
2γ 1−z Λ v0 v − γ 2
408 On Real-Time Calculation of the Rejected Takeoff Distance

We can also obtain explicit formulas to calculate l(t), v(t)

 
1 Λv0
l(t) = ln sh(γt) + ch(γt) , (16.23)
Λ γ
' γ (
v(t) = e−Λl(t) v0 ch(γt) + sh(γt) . (16.24)
Λ
√
For the case ΛG < 0we denote ω = −ΛG, and the solution is

x(t) = C1 sin ωt + C2 cos ωt. (16.25)

Taking into account the initial conditions (16.19), we can obtain explicit
dependency of time t from velocity v
1 Λω (v − v0 )
t = − arctg 2 . (16.26)
ω Λ v0 v + ω 2
We can also calculate l(t), v(t):
 
1 Λv0
l(t) = ln sin ω t + cos ω t , (16.27)
Λ ω
 ω 
v(t) = e−Λl v0 cos ω t − sin ω t . (16.28)
Λ
Thus, at each interval of interpolation of the aircraft parameters with fixed
values, we can analytically express the dependences of distance, time and
speed under given initial conditions. If time is known, we can use (16.23),
(16.24) or (16.27), (16.28), depending on the ΛG sign. If the time is unknown,
but we know the speed at the end of the time interval, then the time can found
from (16.22) or (16.26).

16.8 Computer Simulation

To verify and study the proposed algorithms for estimating the distance
of the interrupted takeoff in real time, a computer simulation system was
created. With this system, we can compare the rejected takeoff distance
obtained using the proposed algorithms with the distance calculated from full
motion model (16.1)–(16.3) using the numerical solution of the differential
equations. The system allows varying numerous aircraft parameters (more
than 50) and initial states. The calculations were performed based on the real
data of the An-124 aircraft. In addition to the above-mentioned parameters,
 16.8 Computer Simulation 409

the temperature, slope of the runway, air density, fuel burn-up rate, number of
running engines, thrust dependence on speed, etc. are also taken into account.
You can set different initial conditions (mass of the aircraft, the speed of
decision-making, after which there is braking, etc.). You can change the time
parameters that affect the process of starting and braking (for example, the
reaction time of the crew, the time characteristics of the transition from full
thrust to reverse, time intervals of change). You can also vary the speed of
transition from full reverse thrust to small gas reverse thrust. With the help
of the system, both the entire path from start to stop (takeoff acceleration and
braking) and the process of only takeoff interruption were investigated (the
speed of decision-making is set and the distance to the moment of stop is
calculated). Different values of engine failure velocity vEF were investigated,
including the values near the decision-making velocity V1 . Actually, for real-
time calculations it is necessary to use the part of the system that emulates
only takeoff interruption. For such calculations, it is possible to use only
8 time intervals with approximation of parameters of system by the fixed
values. It took 10 time intervals to simulate both takeoff and braking, with
sufficient accuracy of emulation. To increase the accuracy of the emulation
of the acceleration part of the rejected takeoff, we can split the time interval
just before the reaching the vEF into several parts (up to 4); in his case we
will use at most 13 time intervals.
The Java programming language was chosen to develop the system. We
used a Java Development Kit version 1.8 (JDK 1.8). Apache Maven version
3.6.3 was used as a project management tool. The user interface (GUI) was
developed using Java Swing.
The main libraries used to develop the system:
Ini4J (Java API for processing configuration files in .ini format for
Windows)–configuration and initial values of system parameters.
Apache Commons Mathematics Library (library of light, independent com-
ponents of mathematics and statistics)-numerical integration of differential
equations describing the motion of the aircraft.
JChart2D (an easy-to-use component written in Java for displaying two-
dimensional tracks in a coordinate system)–visualization of changes in
parameters, speed and coordinates of the aircraft in real time.
Junit, hamcrest, powermock-api-mockito–libraries for the program testing.
A general view of the user interface is shown in Figure 16.2. Using the sliders,
you can change up to 50 parameters; their number, names, as well as the initial
410 On Real-Time Calculation of the Rejected Takeoff Distance

Figure 16.2 User interface

values of these parameters and the ranges of their changes are specified in the
configuration file config.ini. In Figure 16.2, about 30 parameters are available
for changing by sliders, and the remaining 20 are set directly in config.ini.
After changing necessary parameters, we press “Run” button, and the
system performs emulation and equation integration. The results will be
shown as plots, see Figure 16.3. Using checkboxes (left bottom), we can
select the parameters to display. The simulation results are displayed in real
time. For each parameter displayed on the screen, such as speed v, two
values are displayed simultaneously in real time. The bold line (Real) shows
the real value of the parameter obtained from the complete model (16.1)–
(16.3) by numerical integration; thin line (Emulated) represents piecewise
linear approximation of the parameter using the proposed algorithm. The
upper left corner of the screen shows the initial distance, velocity and time.
Top central part shows the distance, velocity and time of the interrupted
takeoff. REAL values are obtained by numerical integration of the complete
model, EMULATED values are obtained using the described algorithm using
two or more (up to 5) piecewise constant approximations, as selected by
 16.9 Emulation Results 411

Figure 16.3 Emulation and integration results

corresponding slider. The upper right corner of the screen shows the values
of the total distance and the time of the interrupted takeoff, at the moment
of full stop with v = 0. REAL values are obtained by numerical integration
of the complete model, EMULATED values are obtained using the described
algorithm using piecewise constant approximations.

16.9 Emulation Results

Figure 16.4 shows the distance, velocity, and thrust from the moment of
engine failure. Such calculations, from the moment of danger, should be used
for calculations in real takeoff. The emulation shows that we can use only 8
time steps while aircraft parameters are changed significantly. The thick line
(Real) shows the aircraft distance and velocity on the runway. These distance
and velocity are calculated using the numerical solution of the differential
equations with the thrust shown on the figure, and with other parameters
that depend on time and velocity (lift, drag, and friction coefficients, mass,
etc.). These parameters are based on values for An–124 aircraft. The thin
412 On Real-Time Calculation of the Rejected Takeoff Distance

Figure 16.4 Rejected takeoff distance, velocity, and thrust

line (Emulated) shows the constant emulated values for the distance, veloc-
ity, and thrust that are calculated using constant estimations of the aircraft
parameters.
Table 16.1 shows the typical values of the distance to the stop after
reaching the decision-making speed, calculated by this algorithm, in com-
parison with the results of numerical integration of the complete model.
From Table 16.1 we can see that the calculation of the interrupted take-
off distance using the proposed algorithm corresponds to the results of
the integration of the motion Equations (16.1)–(16.3). Thus, the piecewise

Table 16.1 Comparison of emulation and integration

Mass, ton 250 276 300 325 350 376
Decision-making velocity, m/s 65.6 66.5 68 68.4 69.5 70.7
Engine failure velocity, m/s 58.4 60.17 62.4 63.3 64.9 66.6
Distance to full stop, m – 1060 1089 1128 1144 1174 1206
emulation using algorithm
Distance to full stop, m – 1057 1085 1125 1141 1172 1204
numerical integration
 16.11 Conclusion 413

Table 16.2 Comparison of emulation and nomograms

Mass, ton 250 275 300 325 350 375
Decision-making velocity, m/s 65.6 66.6 67.5 68.4 69.4 70.7
Rejected takeoff distance, m, 2008 2164 2319 2504 2689 2875
nomogram
Rejected takeoff distance, m – 1962 2127 2305 2499 2710 2982
emulation using algorithm

constant interpolation of engine thrust and other parameters of the aircraft is

legitimate.

16.10 Model Verification

Confirmation of the model adequacy was performed by comparing the results
of mathematical modeling of the motion of the An-124 aircraft during
interrupted takeoff with the results obtained from the nomograms from the
An–124 flight manual. An interrupted takeoff simulation was performed for
several values of decision-making speed V1. and aircraft takeoff mass m. It
was taken into account that the values of time intervals used in the model are
not strictly defined in the regulations, and therefore allow some variation over
the commonly used values.
Table 16.2 shows the results of comparing the lengths of the interrupted
takeoff, calculated using mathematical modeling, and nomograms, with the
same values of the model parameters (nominal thrust, drag coefficients, lift,
friction, etc.).
As we can see from Table 16.2, the simulation allows to obtain results
quite close to the data from the flight nomograms. Thus, the model we used
is adequate and acceptable for studying the influence of hazard factors on the
characteristics of interrupted takeoff.

16.11 Conclusion
The proposed algorithm for calculating the rejected takeoff distance uses step
interpolation of aircraft thrust and other parameters. This interpolation allows
to obtain explicit solutions of the Riccati equation that describes the aircraft
motion. The calculations by explicit formulae on a small number of time
intervals can be performed very quickly, in real time. In any moment of the
takeoff, we can obtain an estimation of the distance to the full stop. Algorithm
takes into account aircraft parameter just in the moment of emergency. With
414 On Real-Time Calculation of the Rejected Takeoff Distance

this algorithm, the accuracy of the rejected takeoff distance estimation is

better comparing to the pre-calculated values. Computer simulation shows
high accuracy of the algorithm. Its simplicity and high speed make it suitable
for using in on-board automatic control systems to provide recommendations
to the crew at the time of an emergency. So the decision about the takeoff
interrupt will be more accurate and safe.

References
[1] Airplane AN-124-100. “Manual for flight” [in Russian], Book 1, 1993.
[2] NLR Air Transport Safety Institute, Report no. NLR-TP-2010-177,
April 2010, https://reports.nlr.nl/xmlui/bitstream/handle/10921/15
8/TP-2010-177.pdf.
[3] “Reducing the risk of runway excursion”. RUNWAY SAFETY INITIA-
TIVE. Flight Safety Foundation, May 2009, https://flightsafety.org/files
/RERR/fsf-runway-excursions-report.pdf.
[4] Kondratenko, Y.P., Kuntsevich, V.M., Chikrii, A.A., Gubarev, V.F.
(Eds). Advanced Control Systems: Theory and Applications. Series in
Automation, Control and Robotics, River Publishers, Gistrup, 2021.
[5] Kondratenko, Y.P., Chikrii, A.A., Gubarev, V.F., Kacprzyk, J. (Eds).
Advanced Control Techniques in Complex Engineering Systems: The-
ory and Applications. Dedicated to Professor Vsevolod M. Kuntsevich.
Studies in Systems, Decision and Control, Vol. 203. Cham: Springer
Nature Switzerland AG, 2019.
[6] A.A. Chikrii. “Control of Moving Objects in Condition of Conflict”.
In: Kuntsevich, V.M. et al. (Eds). Control Systems: Theory and Appli-
cations. Series in Automation, Control and Robotics, River Publishers,
Gistrup, Delft, 2018, pp. 17–42.
[7] A.A. Chikrii, et al. Method of Resolving Functions in the Theory of
Conflict-Controlled Processes. In: Kondratenko, Y.P., Chikrii, A.A.,
Gubarev, V.F., Kacprzyk, J. (Eds) Advanced Control Techniques in
Complex Engineering Systems: Theory and Applications. Dedicated to
Prof. V.M.Kuntsevich. Studies in Systems, Decision and Control, Vol.
203. Cham: Springer Nature Switzerland AG, 2019, pp. 3–33. DOI:
https://doi.org/10.1007/978-3-030-21927-7_1
[8] A.A. Belousov, V.V. Kuleshyn, “Game approach to control of running
start of aircraft on its take off”, J. of Automation and Information
Sciences, 2012, Volume 44, Issue 8, pp. 78–84.
 References 415

[9] A.A. Belousov, V.V. Kuleshyn, V.I. Vyshenskiy, “Real–time algorithm

for calculation of the distance of the interrupted take–off”, J. of Automa-
tion and Information Sciences, 2020, Volume 52, Issue 4, pp. 38–46
[10] V.I. Vyshenskiy, V.V. Kuleshyn., A.A. Belousov , “Computer Simula-
tion of Accelerate-Stop Distance Calculation”, J. of Automation and
Information Sciences, 2020, Volume 52, Issue 11, pp. 72–80
[11] A.A. Chikrii, S.D. Eidelman, “Control game problems for quasi-linear
systems with Riemann-Liouville fractional derivatives”, Cybernetics
and Systems Analysis, 2001, vol. 36, No 6, pp. 836–864.
[12] A.A. Chikrii, G.Ts. Chikrii, “Matrix resolving functions in dynamic
games of approach”, Cybernetics and Systems Analysis, 2014, vol. 50,
No 2, pp. 201–217.
[13] A.A. Chikrii, I.I Matichin, K.A. Chikrii, “Conflict controlled processes
with discontinuous trajectories”, Kibernetika i Sistemnyi Analiz, 2004,
vol. 40, No 6, pp. 15–29 (in Russian)
[14] N.M. Lysenko, “Flight dynamics” [in Russian], VVIA im. prof. N.E.
Zhukovskogo, Moscow, 1967.
[15] V.I. Arnold, “Ordinary differential equations”, Universitext, Springer–
Verlag, Berlin, 2006, 334 pp.
[16] Popov E.P., “Automation regulation and control” [in Russian]. Nauka,
Moscow, 1966. –386 pp.
[17] Shtoda A.V., Morozov F.N., Shiukov A.G., “Systems of control and
regulation of aviation engines” [in Russian], VVIA im. prof. N.E.
Zhukovskogo, Moscow, 1977.
 17
Automated Control Problem for Dynamic
Processes Applied to Cryptocurrency
in Financial Markets

Viktor Romanenko, Yurii Miliavskyi, and Heorhii Kantsedal

Institute of Applied System Analysis of National Technical University

of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”,
37 Peremohy av., Kyiv, 03056, Ukraine
E-mail: [email protected]; [email protected];
[email protected]

Abstract
In this chapter we develop a cognitive map (CM) of the use of cryptocur-
rency in the financial market and describe dynamic model of CM impulse
processes as a difference equation system (Roberts equations) based on
that CM. We chose an external control vector for the CM impulse process
provided by means of CM nodes varying. A closed-loop CM impulse process
control system was implemented. This system includes multidimensional
discrete controller, based on automated control theory methods, that generates
selected control vector and directly affects respective CM nodes by varying
their coordinates. We solved three problems of a discrete controller design for
automated dynamic processes control applied to cryptocurrency in financial
markets. The first problem is unstable cryptocurrency rate stabilization based
on modal CM impulse process control. The second problem is constrained
external and internal disturbances suppression during CM impulse processes
control based on invariant ellipsoid method. The third problem is minimizing
generalized variance of CM nodes coordinates and controls for stabilizing
coordinates at given levels. In this chapter, we developed a system for
identifying CM weighting coefficients based on a recurrent least-squares

417
418 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency

method. We did some performance research for each of the designed discrete
controllers.

Keywords: Cognitive map, linear matrix inequalities, cryptocurrency, state

controller, invariant ellipsoid method, modal control.

17.1 Introduction
To study the dynamic processes applied to cryptocurrency the chapter uses
cognitive modelling, which is one of the most relevant areas of scientific
and practical research of complex systems of different nature now [1–4].
It is based on the notion of a cognitive map (CM), which is a weighted
directed graph [5]. The nodes (nodes) are coordinates (factors) of the com-
plex system and weighted edges (arcs) of the graph describe causal links
between CM nodes. When disturbances affect CM nodes, we can observe
impulse transitional process. Equation (17.1) describes the dynamics of that
process [6]:
!n
Δyi (k + 1) = aij Δyj (k), (17.1)
j=1

where Δyi (k) = yi (k) − yi (k − 1), i = 1, 2, ..., n, aij - weight of an edge

connecting the jth node and the ith one. Equation (17.1) describes the free
motion of the ith node of CM without external control impact. We can write
this equation in vector-matrix form (17.2):

ΔȲ (k + 1) = AΔȲ (k), (17.2)

where A is a weighted adjacency matrix of CM.

In order to implement CM impulse process control based on modern
control theory [7, 8] it is necessary to be able to physically change some
coordinates of CM nodes as control actions. Then we can describe the forced
motion of the CM impulse process under external control as (17.3):

ΔȲ (k + 1) = AΔȲ (k) + BΔU (k), (17.3)

where ΔU (k) = U (k) − U (k − 1)– vector of controls increments with size

m ≤ n. The operator fills the control matrix B and usually uses ones and
zeros.
In [9] stabilization of CM impulse processes in complex systems for m ≤
n based on modal control described in [10] is suggested. We investigated
 17.2 Cognitive Mapping of Cryptocurrency Usage in Financial Markets 419

options of modal control with one and with several controls and carried out
the simulation of a real modal control system of the CM impulse process in
the commercial bank.

17.1.1 Problem Statement

The first problem is to develop a dynamic control model of the CM impulse
process for the application of cryptocurrency in the financial market. The
second problem is the development and research of a modal control algorithm
to stabilize an unstable cryptocurrency rate. The third problem is to develop
a closed-loop CM impulse process control system to suppress constrained
external and internal influences. The control system is based on the invariant
ellipsoids method. In the fourth problem it is necessary to implement a
discrete-time controller for stabilizing the CM nodes coordinates at specified
levels based on the criterion of minimal generalized variance of the nodes
coordinates and the control vector under the influence of perturbations. The
fifth problem is to implement adaptive control of the CM impulse process
with unknown or varying coefficients of the adjacency matrix A. Adaptive
control includes estimating the elements of matrix A during the transient
impulse process and using the estimated values of A to generate the control
vector ΔU (k).

17.2 Cognitive Mapping of Cryptocurrency Usage

in Financial Markets
Cryptocurrency (bitcoin in particular) occupies an intermediate position
between conventional currency and securities. A lot of scientific research has
been devoted to the exploration of this phenomenon [11–14]. Bitcoin has
users who informally create a kind of a “gross domestic product” and there
are companies that accept payments in this cryptocurrency. We can estimate
the number of users by the number of transactions between the so-called
“wallets” and transfers between them (this is publicly available information,
according to which you can see transfers made in bitcoin or other cryptocur-
rency, but “wallets” do not have any information about people). People create
crypto exchanges in order to create contracts to purchase bitcoins and receive
credits. Exchange has a certain number of users and a certain amount of
money comes in every day (trading volume). Several exchanges can function
at the same time (and actually, we have a lot of them now). Each exchange is
itself a common company. Users entrust their money to it and actually support
420 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency

its development. Exchanges form the price of bitcoin. During formation of the
initial price of bitcoin (during mining), it’s worth in the range from 2000 to
3000 dollars, but expectations of bitcoin becoming a conventional payment
system allow it to maintain its price at 9000-10000 dollars. But any financial
news might drastically decrease the price.
There are various perturbations to the processes of using cryptocurrency,
aimed at reducing the level of confidence in the use of cryptocurrency.
Therefore, there are the following risks that can arise in crypto operations.
1. Risk of losing users, which leads to a decrease in the price of bitcoin;
2. Risk associated with inappropriate general expectations of many users
at the same time, which traders create by manipulating the trades on the
exchanges or breaking crypto-related news;
3. The risk of a sharp change in the price of cryptocurrency as a result of
fraud on exchanges like high-frequency trading that usually lies in the
high speed of buying monetary assets before the large operation of an
investor and then selling it to the investor, before the information about
the purchase reaches all exchanges;
4. The risk associated with the lack of guarantee of the safety of capital
invested in the purchase of cryptocurrency, which leads to a certain
hysteria of users during trading on exchanges;
Figure 17.1 shows the CM, which is developed on the basis of causal
relationships.
We introduce the following notations of the CM nodes:
1. rate of cryptocurrency (bitcoin value),
2. trading volume of cryptocurrency,
3. capitalization volume,
4. number of cryptocurrency users,
5. investment volume (interest in bitcoin from institutional investors),
6. volume of speculation in cryptocurrency,
7. supply of cryptocurrency,
8. variance of cryptocurrency rate,
9. demand for cryptocurrency,
10. indirect profit,
11. level of confidence in cryptocurrency,
12. integrated level of risk when using cryptocurrency.
We can measure or compute nodes 1-8. However, we cannot measure nodes
9-12 or it’s hard to measure them.
 17.2 Cognitive Mapping of Cryptocurrency Usage in Financial Markets 421

Figure 17.1 CM cryptocurrency usage in financial markets

As the control actions at the expense of varying the CM node coordinate

resources can be used:
• Varying of cryptocurrency trading volume (Δu1 (k));
• Varying of cryptocurrency capitalization volume (Δu2 (k));
• Varying of investments volume (Δu3 (k))
• Varying of speculation volume (Δu4 (k))
• Varying the offer of cryptocurrency (Δu5 (k)).
The adjacency matrix A has the following form:
⎛ ⎞
1.3 0.25 0.15 0.4 0.6 0.3 0 0 0.2 0 0.4 −0.5
⎜ 0 0 0.15 0.2 0.1 0.7 0.7 −0.4 0.8 0 0.7 −0.5 ⎟
⎜ ⎟
⎜ 0.5 0 0.9 0 0.4 0 0 0 0 0 0.6 0 ⎟
⎜ ⎟
⎜ 0.65 0 0 0.85 0 0 0 0 0 0 0.7 −0.7 ⎟
⎜ ⎟
⎜ 0 0.4 0 0 0.9 0 0 0 0 0.5 0 −0.4 ⎟
⎜ ⎟
⎜ 0 0.7 0 0.4 0 0 0.75 −0.2 0 0 0 −0.5 ⎟
A=⎜ ⎟.
⎜ 0 0 0 0.4 0.4 0 0 0 0.5 0.3 0 0 ⎟
⎜ ⎟
⎜ −0.3 0 0 0 0 0 0 0 0 0 0 0 ⎟
⎜ ⎟
⎜ 0.05 0.8 0 0 0 0.5 0.7 −0.3 0 0 0 −0.4 ⎟
⎜ −0.5 ⎟
⎜ 0 0.7 0.1 0 0 0.5 0 0.1 0.9 0.05 0 ⎟
⎝ 0.5 0 0.5 0 0.3 0 0 −0.3 0.4 0 0 −0.5 ⎠
0 0 0 0 −0.4 0.6 0 0.4 0 −0.2 0 0
422 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency

We use the calculation of the maximum sample variance [15] to determine

the variance of the cryptocurrency rate (node 8). The following algorithm
could define the variance on the interval N T0 :
var {y1 (kT0 ) |y1 ((k − 1) T0 ) , ..., y1 ((k − N) T0 )} =
 1 N
2 (17.4)
= N1 N i=1 y i ((k − i) T 0 ) − N y
j=1 j ((k − j) T 0 )

We chose the sampling time interval N T0 experimentally. We calculate the

maximum sample conditional variance using the sliding duration Nmin T0 ≤
N T0 ≤ Nmax T0 . At each value of N in the specified range with a sam-
pling rate of one sampling period N0 we calculate the sampling conditional
variance and its maximum value is set as:
V8max = sup Nmin T0 ≤ N T0 ≤ Nmax T
[var {y1 (kT0 ) |y1 ((k − 1) T0 ) , ..., y1 ((k − N) T0 )}]
Let us consider the nature of the main perturbations that affect the CM nodes:
1. Informational influences. Since the cryptocurrency environment is quite
open and a large number of cryptocurrency users have no professional
education, the information effects have a significant impact on the fluc-
tuations of the cryptocurrency rate due to the rapid increase in supply
or demand, depending on positive or negative news. A vivid example
of informational impact was Elon Musk’s recommendation to use the
messenger "Signal". In this case, the company’s shares rose by 1300%
in one hour. This kind of perturbation can affect the nodes 3,4,5,7,9,11
of the CM. (At the time of writing this article, Elon Musk made an
announcement about Tesla’s $1.5bln acquisition of bitcoin, but it did not
affect the rate as rapidly as in the first example - only a 40% increase.)
2. Fluctuations in the global economy create uncertainty for institutional
investors (such as large banks), forcing them to look for new areas
for investment. At the same time, the entry of a significant number of
stock market players leads to a sharp increase in the bitcoin rate. Their
behavior in the market is similar to long-term investing and affects nodes
3, 5, 6 CM, which leads to the increase in demand for cryptocurrencies
(node 9). The opposite effect is also possible. If the economy stabilizes,
some investors will move to the stock market, leading to a sharp increase
in the supply of cryptocurrency (node 7).
3. Fluctuations in the price of energy, which is one of the factors of
instability in the world economy. This way the decrease in the price
 17.3 Design of Unstable Cryptocurrency Rate Stabilization System 423

of energy attracts new users to the process of mining and a significant

number of people start to have cryptocurrency. This leads to an increase
in the amount of speculations (node 6) and, naturally, to the growth of
the cryptocurrency rate (node 1).
4. Legislative outrage. For example, the introduction of cryptocurrency as
a unit of payment in Denmark has fuelled interest in cryptocurrency and
dramatically increased the number of users (node 4) and the growth of
speculation (node 6).
5. The mismatch between the level of confidence (node 11) and the volume
of investment (node 5), which is caused by the contradiction between
large and small investors. This mismatch can cause significant fluctu-
ations in the exchange rate of cryptocurrency. Because small investors
use borrowed money, they will be in danger. This leads to a decrease in
the number of cryptocurrency users (node 4).

17.3 Design of Unstable Cryptocurrency Rate Stabilization

System
To stabilize an unstable impulse process in CM, we use the method of modal
control. We design the state controller:

ΔU (k) = −KΔȲ (k). (17.5)

Here the size m of the control vector is smaller than the size n of the vector of
CM nodes coordinates Ȳ (k). This version of the modal control can be found
in [10]. If the pair (A, B) in (17.3) is controllable, the following algorithm
determines the feedback matrix K.
1. We set the desired spectrum of modes λ1 , . . . , λn of the closed-loop
system ΔY (k + 1) = (A − BK)ΔȲ (k), where all λj , j = 1, . . . , n are
different and modulo less than one. Moreover, there are no eigenvalues
of matrix A among λj .
2. We introduce Rj , j = 1, . . . , n – eigenvectors of the state matrix of the
closed system (A-BK), for which this relation is satisfied:

(A − BK)Rj Δ(k) = λj Rj (17.6)

We can write equation in the following form:

(A − λj I)Rj = BKRj = BPj , (17.7)

424 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency

where vector-columns Pj = KRj have dimension m, same as control

vector ΔU .
3. We set an arbitrary matrix P of dimension
) m *× n so that it has full rank
and has no zero columns, Pj = P1 P2 . . . Pn .
4. We calculate the following vectors from equation (17.7):
Rj = (A − λj I)−1 BKRj , j = 1, . . . , n, (17.8)
) *
After that, we form a matrix R = R1 R2 . . . Rn of dimension n × n,
which is non-singular.
5. We calculate the feedback matrix of the modal regulator (17.5) of
dimension m × n:
K = P R−1 , (17.9)
which, by construction, provides the desired set of modes λj of the closed
system. This approach is possible if (A − λj I)−1 exists, so λj is not an
eigenvalue of matrix A. The choice of the matrix P affects the nature of
the controls ΔU (k), but the spectrum of the closed system remains invariant
with respect to P .
We define constraints Δuimin (k) ≤ Δui (k) ≤ Δuimax (k) for the changes
of Δui (k) increments in CM. The constraints are defined based on the
existing resource constraints of the CM nodes, to which the synthesized
Δui (k) is fed. If you do not take this into account, in a real system, the control
actions will be cut off at the level of constraints, so the control system will not
be linear. If information about the maximum possible increments Δyi (k) of
the CM node coordinates is available, the constraints on the value of Δui (k)
in the controller (17.5) can be satisfied by reducing the values of the elements
of the rows of the feedback matrix K. This can be achieved experimentally
by initial selection of vectors Pj to calculate eigenvectors Rj in (17.8).

17.3.1 Experimental Study of the Stabilization System

of Unstable Cryptocurrency Rate
The overall It is easy to see that the system under consideration is unstable,
its modes are equal to 2.352; 1.011; 0.95 ± j0.2921; 0, .5277 ± j0.4123;
−0.6412 ± j0.1454; −0.0753 ± j0.3622; −0.0208 ± j0.0849 (the abso-
lutes of the first two eigenvalues are greater than one). We create the matrix B
depending on the number of control actions. During modelling we set desired
spectrum of eigenvalues λj of the closed-loop control system state matrix
A − BK as following: −0.6; −0.5; −0.4; −0.3; −0.2; −0.1; 0.1; 0.2;
0.3; 0.4; 0.5; 0.6, with absolutes that are less than one.
 17.4 The Problem of Constrained Internal and External Disturbances 425

Consider the case when control actions Δu1 (k), Δu2 (k), Δu3 (k),
Δu4 (k), and Δu5 (k) are applied to the following nodes of CM respectively:
2 - trading volume, 3 - capitalization volume, 5 - investment volume, 6 -
speculation volume, 7 - supply of cryptocurrency. In this case, the matrix B
has the following form:
⎛ ⎞
0 1 0 0 0 0 0 0 0 0 0 0
⎜ 0 0 1 0 0 0 0 0 0 0 0 0 ⎟
⎜ ⎟
B =⎜
T ⎟
⎜ 0 0 0 0 1 0 0 0 0 0 0 0 ⎟.
⎝ 0 0 0 0 0 1 0 0 0 0 0 0 ⎠
0 0 0 0 0 0 1 0 0 0 0 0

We form the matrix P for the design of the modal controller as follows:
⎛ ⎞
1 1 1 1 1 1 1 1 1 1 1 1
⎜ 1
⎜ 0 0 1 1 0 0 1 0 1 1 0 ⎟ ⎟
P =⎜ ⎜ 0 0 1 0 0 1 0 0 0 0 1 1 ⎟ ⎟.
⎝ 0 1 1 1 1 1 0 1 0 1 1 1 ⎠
1 1 0 0 0 0 0 1 1 1 0 0

Based on methods (17.6), (17.7), (17.8), we define the feedback matrix of the
modal controller (17.9).
Figure 17.2 shows the results of modelling the CM system of equations
under five control actions, given that initial negative impulse “−1” affects
node 11 “confidence level”. The graphs show transients of the CM node
coordinates yi , i = 1, . . . , n, coordinate increments Δyi (k), and controls
increments Δu1 (k), Δu2 (k), Δu3 (k), Δu4 (k), and Δu5 (k).
Based on the analysis of the graphs, we can state the following patterns
of changes in the CM node coordinates and their increments during the
transition process:
• The coordinates: y7 (supply of cryptocurrency) and y10 (indirect profit)
increased;
• The following coordinates decreased: y1 (cryptocurrency rate), y2 (trad-
ing volume) and y4 (number of users), y8 (rate variance), y12 (risk
level).
The node coordinate increments Δyi (k) , i = 1, . . . , 12 during the sta-
bilization transient converge to zero, i.e. we suppress all perturbations.
The increments of all control actions Δui (k) converge to zero after the
coordinates of nodes yi are stabilized.
426 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency

y (k ) y (k ) y (k ) y (k )
1 2 3 4
5 6 5 5

4 4 4

3 2 3

2 0 -5 2
0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20

y (k ) y (k ) y (k ) y (k )
5 6 7 8
8 8 12 6

10
7 7 4

6 6 2
6

5 5 4 0
0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20

y (k ) y (k ) y (k ) y (k )
9 10 11 12
10 10 5 5

8 8 4

4.5

6 6 3

4 4 4 2
0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20

y (k ) y (k ) y (k ) y (k )
1 2 3 4
0 5 0 1

-1 0.5
-0.5

0 -2 0

-1
-3 -0.5

-1.5 -5 -4 -1
0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20

y (k ) y (k ) y (k ) y (k )
5 6 7 8
1.5 4 6 2

1 4 0
2

0.5 2 -2

0
0 0 -4

-0.5 -2 -2 -6
0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20

y (k ) y (k ) y (k ) y (k )
9 10 11 12
2 2 1 1

0.5
1 1 0

0 0 -1
-0.5

-1 -1 -1 -2
0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20

Figure 17.2 Continued

 17.4 The Problem of Constrained Internal and External Disturbances 427
u (k ) u (k )
1 2
5 5

0
0
-5

-10 -5
0 5 10 15 20 0 5 10 15 20

u (k ) u (k )
3 4
2 5

1
0
0

-1 -5
0 5 10 15 20 0 5 10 15 20

u (k )
5
5

-5
0 5 10 15 20

Figure 17.2 Results of modelling for the CM

17.4 The Problem of Constrained Internal and External

Disturbances Suppression in Control
of the Cryptocurrency CM Impulse Process
In [16, 17] theoretical positions on the suppression of arbitrary constrained
external disturbances in terms of invariant ellipsoids based on the design of
a static state feedback, which minimizes the size of the invariant ellipsoid of
the dynamical system, are presented. In this case, we implemented a robust
control, the analysis and synthesis problem of which is reduced to equivalent
conditions in the form of linear matrix inequalities (LMI), solved numerically
on the basis of semi definite programming.
In [18] we solve the problem of suppression of constrained external
disturbances based on the invariant ellipsoids approach in the implementation
of a closed-loop control system of impulse processes in CM of complex
systems. The general model of the dynamics of impulse processes (17.2) is
decomposed into two mutually related systems of difference equations:
ΔX (k + 1) = A1 ΔX (k) + DΔZ (k) ; (17.10)
ΔZ (k + 1) = CΔZ (k) + ΨΔX (k) . (17.11)
Here X is the vector of measurable coordinates of CM nodes; Z is the vector
of unmeasurable or hardly measurable coordinates. The matrices A1 , D,
428 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency

C, Ψ are compiled from the coefficients of the matrix of the initial model
(17.2) of the CM impulse process. The matrices D, Ψ show the relationships
between the first (17.10) and the second (17.11) parts of the initial CM (17.2).
The values of non-measurable coordinates ΔZ(k) are taken into account as
external constrained perturbations with unknown probability characteristics
in the first system of equations (17.10) of the CM model compiled for impulse
processes with measurable coordinates X.
We synthesized a control vector to suppress constrained perturbations
ΔZ(k) by implementing static state feedback:

ΔU (k) = −Kp ΔX(k), (17.12)

which acts directly on the measured coordinates of the nodes X of the first
control system according to the state equation:

ΔX(k + 1) = A1 ΔX(k) + BΔU (k) + DΔZ(k) (17.13)

The control is performed by changing the resources of the CM nodes, which

are influenced by the vector ΔU (k).
In this chapter, the change in the weight coefficients ΔA1 (k) with respect
to the known basic values of the estimated matrix Â1 (k) is proposed to be
considered as the internal perturbations in the CM impulse process model
(17.10). For this purpose, we propose to consider the initial model (17.10) in
the form:

ΔX(k + 1) = A1 ΔX(k) + ΔA1 ΔX(k) + DΔZ(k), (17.14)

where ΔA1 = A1 − A1var (k) is the change in the adjacency matrix of CM

(17.10) during the sampling period, where A1var (k) is the real unknown value
of the matrix A1 , which changes as the complex system evolves.
Let us denote the increment of internal perturbations in (17.14) as
ΔA1 (k)Δy(k) = Δw(k). Then the equation of the uncontrolled impulse
process (17.14) will take the form:
 
Δw(k)
ΔX(k + 1) = A1 ΔX(k) + ( I1 D ) , (17.15)
ΔZ(k)

in which the vectors and matrices have the following dimensions: dim ΔX =
n; dim ΔZ = p;dim Δw = n;A1 (n × n); D(n × p), I is a unit matrix of
dimension n × n. We assume that the internal and external perturbations are
 17.4 The Problem of Constrained Internal and External Disturbances 429

jointly constrained by the norm l∞ , so that:

    1/
 Δw(k)  8 9 Δw(k) 2
  T T ≤1
 ΔZ(k)  = sup Δw̄ (k) ΔZ̄ (k)
ΔZ(k)
∞
(17.16)
We propose invariant ellipsoids on state variables to describe the character-
istic of the effect of perturbations of the type (17.16) on the trajectory of a
dynamic discrete system (17.15), as in [16, 17]:
6 7
εΔX = ΔX(k) ∈ Rn : ΔX̄ T P −1 ΔX ≤ 1 , P > 0, (17.17)

if from ΔX(0) ∈ εΔX the condition ΔX(k) ∈ εΔX follows for all discrete
moments of time k = 1, 2, 3, . . . . Then the matrix P is called the matrix of
the ellipsoid εΔX .
Let us prove the invariance condition of the ellipsoid (17.17) under
perturbations (17.16). For this purpose, using the method [17], we introduce
the quadratic Lyapunov function V (X(k)) = ΔX̄ T QΔX̄ at Q > 0, con-
structed on the solutions of the system (17.15). To ensure that the trajectories
ΔX(k) 6 of the system (17.15) do not go 7beyond the boundary of the ellipsoid
εΔX = ΔX(k) ∈ Rn : V (X(k)) ≤ 1 , it is required that V (ΔX(k)) ≤ 1
at V (X(k)) ≤ 1, so we have:

ΔX̄ T (k + 1)QΔX̄(k + 1) =
  
8 9 I 1
= ΔX̄ T A1 T + Δw̄T (k) ΔZ̄ T (k) Q
DT
  
) * Δw(k)
A1 ΔX̄(k) + I1 D ≤1
ΔZ(k)

By multiplication, we get:

ΔX̄ T (k + 1)QΔX̄(k + 1) =
8 8 9 9
= ΔX̄ T Δw̄T (k) ΔZ̄ T (k)
⎡ ⎤
A1 T QA1 A1 T Q (I1 D )
T ⎡ ⎤
⎢ ⎥  ΔX(k) 
⎢     ⎥⎣ Δw(k) ⎦
⎣ I1 I1 T ⎦
QA1 Q (I1 D ) ΔZ(k)
DT DT
430 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency
⎡ ⎤
 ΔX(k) 
Let us apply the S-procedures [17, 19]. Let S = ⎣ Δw(k) ⎦. Then the
ΔZ(k)
quadratic forms we can represent as follows:
T
f0 (S) = S M0 S =
8 8 9 9
= ΔX̄ T Δw̄T (k) ΔZ̄ T (k)
⎡ T ⎤⎡ ⎤ ;
A1 T QA TQ
 A1  1(I D )  ΔX(k) 
⎣  I1  1
I1 T ⎦⎣ Δw(k) ⎦ ≤ 1
QA1 Q (I1 D )
DT DT ΔZ(k)
8 8 9 9
f1 (S) = S̄ T M1 S = ΔX̄ T Δw̄T (k) ΔZ̄ T (k)
⎡ ⎤
  ΔX(k)
Q 0 ⎣  
Δw(k) ⎦ ≤ 1;
0 0
ΔZ(k)
8 8 9 9
f2 (S) = S̄ T M2 S = ΔX̄ T Δw̄T (k) ΔZ̄ T (k)
⎡ ⎤
  ΔX(k) 
Q 0 ⎣ 
Δw(k) ⎦ ≤ 1,
0 I2
ΔZ(k)

× (n + p). According to the

where I2 is the unit matrix of dimension (n + p)
statement of S-procedure [17] we have M0 ≤ 2i=1 τi mi , that is:
⎡ ⎤
A1 T QA A1 T Q (I1 D )
T    
⎣   1   Q 0 0 0
I1 I1 T ⎦ ≤ τ1 +τ2
T QA1 T Q (I1 D ) 0 0 0 I2
D D
or ⎡ ⎤
T
 A1 T QA
 1  A1 T Q
 (I1 D )
⎣ I1 I1 T ⎦≤0 (17.18)
T QA1 T Q (I1 D )
D D
Using Schur’s formula, inequality (17.18) will take the following form:
  −1
) * I1 ) *
A1 T QA1 − τ1 Q ≤ A1 T Q I1 QT Q I 1 Q T − τ I
2 2
DT
 
I1
QA1
DT
 17.4 The Problem of Constrained Internal and External Disturbances 431

After performing elementary transformations, we convert this inequality

to the following form:
τ1 Q ≥ A1 T
   −1   
) * I1 ) * I1
Q − Q I1 Q T Q I1 Q T − τ2 I2 Q A1
DT DT

If τ2 = 1 − τ1 then we get:
τ1 Q ≥ A1 T
    −1   
I1 ) * I1
Q + Q(I1 QT ) (1 − τ1 )I2 − Q I1 D Q A1
DT DT
(17.19)
According to the lemma of matrix inversion [20], we will have:
     
) * I1 ) * −1 I1
Q + Q I1 Q T (1 − τ1 )I2 − Q I1 D Q
DT DT
  −1
) * I1
== Q−1 − (1 − τ1 )−1 I1 D
DT
Then expression (17.19) we can write as follows:
  −1
−1 −1
) * I1
τ1 Q = A1 Q − (1 − τ1 )
T
I1 D A1
DT

Let us perform an elementary transformation in condition P = Q−1 :

    
−1 −1 −1 −1
) * I1 ) T *−1 −1
τ1 P ≥ A1 Q − (1 − τ1 ) I1 D A1
DT
After inverting the left and right parts, we get:
  
P −1 −1
) * I1 ) T *−1
≥ A1 P − (1 − τ1 ) I1 D T A1
τ1 D

Multiply from the left by A1 , and then from the right by A1 T , and
renominate τ1 = α. Then the linear matrix inequality will take the final form:
1 I1 + DDT
A1 P A1 T − P + ≤0 (17.20)
α (1 − α)
432 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency

17.4.1 Algorithm for the State Controller Design for the CM

Impulse Process
The equation of state of the controlled impulse process of CM (17.15) under
additional internal perturbation takes the form:
 
) * Δw(k)
ΔX(k + 1) = A1 ΔX(k) + BΔU (k) + I1 D (17.21)
ΔZ(k)
When the state controller (17.12) is applied, the equation of the closed-loop
CM impulse process control system is written as follows:
 
) * Δw(k)
ΔX(k + 1) = (A1 − BKp )ΔX(k) + I1 D (17.22)
ΔZ(k)
) *
It is assumed that the pair A1 B , in the model (17.21) is controllable.
Then the LMI (17.20) for the closed system takes the form:

1 I1 + DDT
(A1 − BKp )P (A1 − BKp )T − P + ≤0 (17.23)
α (1 − α)
We consider the minimization of the trace of the ellipsoid matrix (17.17)
as the optimality criterion for the synthesis of the regulator (17.12) in this
chapter:
trP (α) → min, α∗ ≤ α < 1, (17.24)
This ensures minimization of the size ofthe invariant
 ellipsoid (17.17) with
Δw(k)
the largest suppression of perturbations , which are constrained
ΔZ(k)
only by the maximum range (17.16). After multiplying the factors in the
inequality (17.23), we obtain:
1
(A1 P A1 T − BKp P A1 T − A1 P Kp T B T + BKp P Kp T B T )
α
I1 + DDT
−P + ≤0 (17.25)
(1 − α)
Inequality (17.25) is nonlinear with respect to P and Kp , which need to be
optimized. In [17] a linearization by replacement L = Kp P and introduction
of an additional constraint is proposed:
 
R L
≥ 0, (17.26)
LT P
 17.4 The Problem of Constrained Internal and External Disturbances 433

where P = RT . This inequality is equivalent to R ≥ LP −1 LT = Kp P Kp T

according to the Schur’s formula at P > 0. Then to fulfil inequality (17.25)
it is sufficient that:
1 I1 + DDT
(A1 P A1 T −BLA1 T −A1 LT B T +BRB T )−P + ≤ 0 (17.27)
α (1 − α)
Minimization of criterion (17.24) under constraints (17.26), (17.27) is per-
formed on variables P, L, R by semidefinite programming method by using
Matlab-based SeDuMi Toolbox.
Then the matrix K̂p of the optimal state regulator (17.12) we define as:

K̂p = L̂P̂ −1 (17.28)

with the estimated values of α̂, P̂ , L̂, R̂, providing minimization of criterion
(17.24) under constraints (17.26), (17.27).

17.4.2 Experimental Study of the System of Constrained Internal

and External Disturbances Suppression in the
Cryptocurrency CM Impulse Process Control
Consider a stable version of the cognitive map of cryptocurrency shown in
Figure 17.1. Nodes coordinates 1, 2, 3, 4, 5, 6, 7, 8 are used as measurable
and computable coordinates X̄ in model (10). The group of coordinates Z
that are not measured or difficult to measure includes demand for cryp-
tocurrency (node 9), indirect profit (node 10), confidence level (node 11)
and integral risk level (node 12). After decomposing model (17.3) with
updated coefficients into (17.10) and (17.11), the matrices A1 , D are the
following:

⎛ ⎞
0.3 0.1 0.15 0.2 0.2 0.3 0 0
⎜
⎜ 0 0 0.1 0.1 0.1 0.35 0.35 0 ⎟ ⎟
⎜
⎜ 0.25 0 0.5 0 0.1 0 0 0 ⎟ ⎟
⎜ 0.4 0 0 0.3 0 0 0 0 ⎟ ⎟,
A1 = ⎜
⎜
⎜ 0 0.4 0 0 0.4 0 0 0 ⎟ ⎟
⎜
⎜ 0 0.5 0 0.2 0 0 0.6 −0.2 ⎟
⎟
⎝ 0 0 0 0.35 0.3 0 0 0 ⎠
−0.3 0 0 0 0 0 0 0
434 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency
⎛ ⎞
0.2 0 0.3 −0.3
⎜ −0.3 0 0.3 −0.2 ⎟
⎜ ⎟
⎜ 0 0 0.5 0 ⎟
⎜ ⎟
⎜ 0 0 0.2 −0.2 ⎟
D=⎜ ⎟.
⎜ 0 0.4 0 −0.3 ⎟
⎜ ⎟
⎜ 0 0 0 −0.4 ⎟
⎜ ⎟
⎝ 0.5 0.2 0 0 ⎠
0 0 0 0
The control matrix B has the form given in the example of paragraph 3 (cut
to 8 rows).
We consider the step impacts with the amplitude of 1, acting at the initial
moment of time on the two unmeasured nodes 10 and 11 in the direction
of reducing the indirect profit (node 10) and confidence level (node 11) for
modelling the dynamics of the closed-loop control system of the impulse
process of this CM cryptocurrency using the proposed method, as external
disturbances. The internal perturbations in the simulation are formed as fol-
lows: the value of non-zero matrix coefficients vary at each sampling period
by the formula A1var (k) = A1 ξ(k), where ξ(k) is a normally distributed
random variable (Gaussian white noise). Only the values of A1 are used for
control, while A1var remains unknown. The initial levels of all coordinates of
the CM nodes are assumed zero for convenience.
The charts of the transients of the CM node coordinates, their increments
and controls are shown in Figure 17.3.

17.5 The Problem of Stabilizing the Coordinates of

Cryptocurrency CM Nodes at Given Levels Based on
Varying of Edge Weights and Nodes Coordinates
Provided that limk→∞ ΔXi (k) = 0, in order to stabilize the coordinates of
measured CM nodes Xi at specified levels Gi at each sampling period it is
necessary to form the control actions, which according to the synthesized
control law affect directly the CM nodes Xi in the closed-loop control system.
However, in practice, difficulties arise in controlling impulse processes in
CM due to small resources for varying the coordinates of CM nodes when
implementing control actions. Usually if the CM dimension is large, there
are not enough nodes available for variation, which leads to a significant
difference between the dimension n of the vector of output-controlled coordi-
nates Xi and the dimension m of the vector of control actions ui . Therefore,
17.5 The Problem of Stabilizing the Coordinates of Cryptocurrency CM Nodes 435
y y y y
1 2 3 4
5 5.5 5 5

5 4 4

0 4.5 3 3

4 2 2

-5 3.5 1 1
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50

y y y y
5 6 7 8
5 7 5 8

4.5 4
6 7

4 3

5 6
3.5 2

3 4 1 5
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50

y y y y
9 10 11 12
5 5 5 9

4 4 8

3 3 0 7

2 2 6

1 1 -5 5
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50

y y y y
1 2 3 4
0.5 0.5 0.5 0.5

0 0 0

-0.5 -0.5 -0.5

-1 -0.5 -1 -1
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50

y y y y
5 6 7 8
0.2 0.5 0.5 0.4

0
0 0.2

-0.2 0

-0.5 0
-0.4

-0.6 -0.5 -1 -0.2

0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50

y y y y
9 10 11 12
0.2 0.5 0 0.4

0
0 0.2

-0.2 -0.5

-0.5 0
-0.4

-0.6 -1 -1 -0.2
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50

Figure 17.3 Continued

we decided [21] to apply also another approach, specifically, to generate an

external control vector by simultaneously varying the resources of the CM
node coordinates and the weight coefficients of the adjacency matrix in the
436 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency
u u
1 2
0.4 1

0.2 0.5

0 0

-0.2 -0.5
0 10 20 30 40 50 0 10 20 30 40 50

u u
3 4
0.4 1

0.2 0.5

0 0

-0.2 -0.5
0 10 20 30 40 50 0 10 20 30 40 50

u
5
0.4

0.2

-0.2
0 10 20 30 40 50

Figure 17.3 CM coordinates and controls changes

closed-loop control system. In this case, the general mathematical model of

the controlled impulse CM process (17.13) in vector-matrix form in full nodes
coordinates is represented in the form:
 
−1
8 9 ΔU (k)
X(k + 1) = (I + A1 − A1 q )X(k) + B L(k) + ξ(k),
Δa(k)
(17.29)
where the matrix L(k) is composed of the measured CM coordinates Xμ (k),
which through edges with varying weight coefficients Δaiμl (k) will be used
as the control action; q −1 - backshift operator for one sampling period;
ξ(k) = DΔz(k) - vector of uncontrollable perturbations, according to (13).
Let us formulate the rules for the formation of the vectors of increments
ΔU (k) and Δa(k), as well as the matrix B, L(k) in the model (17.29).
1. Only one control action can influence on each node of CM Xi (k), i =
1, . . . , n by varying the resources of this node Δui or by varying the
weighting coefficient Δaiμl . Therefore, we can guarantee the autonomy
of the control system.
2. The vector of increments of weight coefficients Δa(k) of dimension
p at p < n contains only nonzero elements Δaiμl = 0, where μ is
the number of the CM node that affects the ith node Xi through the
17.5 The Problem of Stabilizing the Coordinates of Cryptocurrency CM Nodes 437

edge with the coefficient aiμ . l = 1, 2, . . . , p are the ordinal numbers

of this coefficient in the vector Δa(k). If the weight coefficient aiμ for
a CM edge entering a node Xi cannot be varied, then the increment
Δaiμl (k) = 0 and in the vector Δa(k) are not considered.
3. The matrix B in (17.29) is created by a designer of the CM impulse
process control system. This matrix is designed to organize scaling and
switching of the synthesized control action Δui , i = 1, 2, . . . , m. The
dimension of the matrix B is equal to n × m, where n is the dimension
of the vector X, and m is the dimension of the vector Δu(k). As a rule,
the elements of B are ones and zeros. Elements biμ = 1 if the ith control
action affects the μth node of the matrix. Thus in each row of the matrix
B only one element can be equal to one, and the remaining elements will
be equal to zero.
4. The matrix L(k) with dimension n × p contains only one element in
each i row equal to the measured coordinate Xμ (k), which is placed in
the lth row, where l is the ordinal number of the increment Δaiμl (k) in
the vector Δa(k), and μ is the number of the CM node, through which
the increment Δaiμl (k) controls the CM node Xi .
5. All elements in the ith row of the matrix L(k) are equal to zero if the ith
node Xi does not include a CM edge with a varying weight coefficient,
i.e. Δaiμ = 0.
Thus, the total number of controls in a CM can be equal to m + p ≤ n.

17.5.1 Design of a Discrete Controller

 
ΔU (k)
We implement the synthesis of the combined vector of control
Δa(k)
based on minimization of the following quadratic criterion of optimality:
  T
8 9T 8 9 ΔU (k)
J(k + 1) = E X(k + 1) − G X(k + 1) − G +
Δa(k)
  
R1 0 ΔU (k)
,
0 R2 Δa(k)

where G is the reference vector to stabilize the coordinates of the CM nodes

at the given levels; R1 , R2 are the weight diagonal positive definite matrices;
E is the expectation operator. Taking into account Equation (17.29), let us
perform minimization of this criterion with respect to the combined control
vector.
438 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency
 
∂J(k+1) BT
  == 2
ΔU (k) LT (k)
∂
Δa(k)   
8 −1
9 8 9 ΔU (k)
I + A 1 − A1 q X(k) + B L(k) + ξ(k) − G +
Δa(k)
  
R1 0 ΔU (k)
+2 =0
0 R2 Δa(k)

From where we get the law of the combined optimal control:

   
ΔU (k) BT 8 9
= T B L(k) +
Δa(k) L (k)
 −1  
R1 0 BT 68 9 7
+ I + A1 − A1 q −1 X(k) + ξ(k) − G
0 R2 LT (k)
(17.30)
Based on equations (17.29), (17.30), the equation of dynamics of the
closed system of the CM impulse process equation can be derived.
  
8 9 BT B B T L(k)
X(k + 1) = I − B L(k)
LT (k)B LT (k)L(k)
  −1  
R1 0 BT 8 9
+ × I + A1 − A1 q −1 X(k)+
0 R2 LT (k)
   −1
BT B B T L(k) R1 0
+ [B L(k)] + G+
LT (k)B LT (k)L(k) 0 R2
  
8 9 BT B B T L(k)
+ I − B L(k)
LT (k)B LT (k)L(k)
  −1  
R1 0 BT
+ ξ(k) (17.31)
0 R2 LT (k)

The stability of the closed-loop control system (17.31) is determined by

the eigenvalues varying in time of the nonlinear matrix expression:
   −1  
BT B B T L(k) R1 0 BT
I − [B L(k)] +
LT (k)B LT (k)L(k) 0 R2 LT (k)
(17.32)
17.5 The Problem of Stabilizing the Coordinates of Cryptocurrency CM Nodes 439
 
BT B B T L(k)
In [21] the proof is given: the matrix +
  L (k)B LT (k)L(k)
T

R1 0
in expression (17.32) is diagonal and positively determined,
0 R2
since on the basis of the rule of formation of matrices B and L(k)
it turns out that B T L(k) = 0, LT (k)B = 0. As a result, it is
shown
 in [21] that the eigenvalues of the diagonal matrix  (17.32) I −
B(B T B + R1 )−1 B T 0
will be mod-
0 L(k)(LT (k)L(k) + R2 )−1 LT (k)
ulo less than unity. As a result, the closed-loop system (17.31) will be
stable.

17.5.2 Experimental Studies of the Stabilization

of Cryptocurrency CM Nodes Coordinates
at Given Levels
We use the coordinates of nodes 1, 2, 3, 4, 5, 6, 7, 8 of the CM shown in
Figure 17.1 as measurable coordinates X in the model (17.29). We include
nodes 9, 10, 11, and 12 in the group of non-measurable or hard-to-measure
coordinates Z. The values of the matrices A1 , D are given above. The
control vector by varying the node coordinate resources is equal to Δu(k) =
8 9T
Δu1 (k) Δu2 (k) Δu3 (k) Δu4 (k) , which are fed respectively to
the following nodes of CM: 2 - trading volume; 5 - investment volume; 6
- speculation volume and 7 - cryptocurrency supply. Then the control matrix
B will be formed by the system designer as follows:
⎛ ⎞
0 1 0 0 0 0 0 0
⎜ 0 0 0 0 1 0 0 0 ⎟
BT = ⎜ ⎝ 0 0 0 0 0 1 0 0 ⎠.
⎟

0 0 0 0 0 0 1 0

As a result, the product of matrices B T B will be the unit matrix 4 × 4. The

second control vector based on the increments of the weight coefficients Δa
8 9T
will have the formΔa = Δa15 Δa 35 . Then the control matrix L(k)
of dimensionality 8 × 2 is composed as follows:
1 2
x5 (k) 0 0 0 0 0 0 0
T
L (k) = .
0 0 x5 (k) 0 0 0 0 0
440 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency

y (k ) y (k ) y (k ) y (k )
1 2 3 4
8 6 8 8

7 5.5 7 7

6 5 6 6

5 4.5 5 5
0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40

y (k ) y (k ) y (k ) y (k )
5 6 7 8
5.5 5.5 6 5

5.5
5 5 4.5
5

4.5 4.5 4.5 4

0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40

y (k ) y (k ) y (k ) y (k )
9 10 11 12
7 6.5 8 5

6 6 7 4.8

5 5.5 6 4.6

4 5 5 4.4
0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40

Figure 17.4 Results of applying the control actions

As a result the product of matrices LT (k)L(k) will be a diagonal matrix 2×2:

 2 
T x5 (k) 0
L (k)L(k) =
0 x5 2 (k)

Let the initial values of CM node coordinates, which are measured on a

10-point scale, be at the average level of 5. Suppose that it is necessary to
transfer the coordinates of the nodes x1 (cryptocurrency rate), x3 (capitaliza-
tion volume) to a higher level 7. Applying the control law (17.30), we obtain
the dynamics of the impulse process control presented in Figure 17.4.

17.6 Design of a System for Identifying CM Weighting

Coefficients Based on Recurrent Least Squares
Method
The model of the controlled impulse process (17.3) of the input-output CM
type can be represented as:

(I − A1 q −1 )ΔY (k) = Bq −1 ΔU (k) (17.33)

Weighting coefficients of the incidence adjacency matrix A are usually

determined by applying expert evaluations. In the process of evolving of the
 17.7 Conclusion 441

cryptocurrency application system, these coefficients may change over time,

depending on changes in the influence of one of the CM nodes, i.e. both
control and evaluation of the adjacency matrix A parameters (coefficients)
must be performed simultaneously.
Let us describe the control (17.33) coordinate-wise:
!
n
Δyi (k) = aij Δyj (k − 1) + bi Δui (k − 1) + ξi (k) (17.34)
j=1

It is assumed that the perturbations ξi (k), caused by inaccurate knowledge

of the model coefficients, are white noise. It should be taken into account
that the structure of the matrix A is known and some of the coefficients aij
are obviously equal to zero (in those cases when there are no connections
between the corresponding CM nodes). Also, when the dimensionality of the
control vector m < n, then in (17.34) some control actions Δui (k − 1) will
be equal to zero.
Let the nonzero elements in the ith row of the matrix A be the coef-
ficients aij1 . . . aijP1 . Denote X̄iT (k) = [ΔYj1 (k − 1), . . . , ΔYjPi (k − 1)],
Θi = [aij1 . . . aijP1 ]T . Then expression (17.34) can be written as follows:
Δyi (k) − bi Δui (k − 1) = X̄iT (k)Θi + ξi (k) (17.35)
ˆ (k). To estimate
The current estimate of the vector Θi is denoted by Θ̄ i
the weight coefficients of the matrix A we apply the recurrent least squares
method [22–25]:
 
ˆ (k) = Θ̄
Θ̄ ˆ (k − 1) + K (k) Δy (k) − b Δu (k − 1) − X̄ T (k)Θ̄ ˆ (k − 1)
i i i i i i i i

1
Ki (k) = Pi (k − 1)X i (k)X̄iT (k)Pi (k − 1)
1+ X̄iT (k)Pi (k − 1)X i (k)
(17.36)
1
Pi (k) = Pi (k − 1) −
1 + X̄iT (k)Pi (k − 1)X i (k)
Pi (k − 1)X i (k)X̄iT (k)Pi (k − 1).
The recurrent procedure (17.36) should be performed for each node of CM
Δyi (k), i = 1, 2, . . . , n at each sampling step. We use the obtained estimates
ˆ (k) as the values of the adjacency matrix A coefficients at the current step
Θ̄ i
of the control algorithm. For parametric identification of the adjacency matrix
A, we can also apply non-recurrent identification methods, outlined in [8, 26].
442 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency

17.7 Conclusion
This chapter considers the cryptocurrency CM and solves three problems of
its application in the financial markets. The first problem covers the case of
instability when the cryptocurrency rate is unstable and decision-makers want
to stabilize it. For this case control actions are designed based on the modal
controller, which help to stabilize the system and assign desired values to the
closed-loop system poles. The second problem considers the cryptocurrency
system affected by unmeasurable internal and external disturbances, and the
only thing we know about these disturbances is that they are constrained.
For this case controls are designed based on the modified invariant ellipsoids
method which ensures that the system’s “trajectory” is limited inside of the
minimal size ellipsoid. The third problem discussed in the chapter relates to
the case when it is required to set some of the CM nodes at the new levels.
This problem is solved based on minimization of the generalized variance
criterion. Additional feature of this approach is that we may not only change
some resources associated with CM nodes but also change how one node
affects another one, i.e. to vary CM edges weights. As a result of solving
all three problems decision-makers obtain a sequence of control actions
which they can apply to the real cryptocurrency market. All the results were
mathematically simulated and the effectiveness of the suggested approaches
was proven.

References
[1] E. Papageorgiou, J. Salmeron, “A review of fuzzy cognitive maps
research during the last decade”, IEEE transactions on fuzzy systems,
vol. 21, no. 1, 2013, pp. 66–79.
[2] G. Gorelova, N. Pankratova (Eds.). Innovational development of socio-
economic systems based on methodologies of foresight and cognitive
modeling, Kyiv, Naukova dumka, 2015, 464 p. (In Russian)
[3] C. Stylios, P.P. Groumpos, “Modeling complex systems using fuzzy
cognitive maps”, IEEE Transactions on Systems, Man, and Cybernetics-
Part A: Systems and Humans, 34((1), 155–162, 2004.
[4] J. Aguilar, “A survey about fuzzy cognitive maps papers”, International
Journal of Computational Cognition, 3, N 2, P. 27–33, 2005.
[5] R. Axelrod, “The structure of decision: Cognitive maps of political
elites”, Princeton University Press, 1976.
 References 443

[6] F. Roberts, “Discrete mathematical models with applications to social

biological and environmental problems”, Englewood Cliffs: Prentice-
Hall, 1976.
[7] V.M. Kuntsevich, et al (Eds). Control Systems: Theory and Applica-
tions. Series in Automation, Control and Robotics, River Publishers,
Gistrup, Delft, 2018.
[8] Y.P. Kondratenko, V.M. Kuntsevich, A.A. Chikrii, V.F. Gubarev,
(Eds). Advanced Control Systems: Theory and Applications. Series in
Automation, Control and Robotics, River Publishers, Gistrup, 2021.
[9] V.D. Romanenko, Yu.L. Milyavsky. “Stabilization of impulse processes
in cognitive maps of complex systems on the basis of modal state reg-
ulators”, Cybernetics and Computer Engineering, vol. 179, pp. 43–55,
2015. (In Russian)
[10] Methods of Classical and Modern Automatic Control Theory: book in
3 Volumes. – T.2: Synthesis of Controllers and Optimization Theory of
Automatic Control Systems / Ed. by N.D. Egupov. Moscow: Bauman
Moscow State Technical University, 2000, 736 p. (In Russian)
[11] Z. Dvulit, Kh. Peredalo, R. Tylipska, R. Terno, R. Stubel, “Cryptocur-
rency: state and trends of development”, Economy and state, 2019, N.
1, P. 10–14. (In Ukrainian)
[12] R. Orastean, S. Marginean, R. Sava, “Bitcoin in the scientific literature
– A bibliometric study”, Studies in Business and Economics, 14(3), p.
160–174, 2019.
[13] A.F. Bariviera, M.J. Basgall, W. Hasperue, M. Naiouf, “Some stylized
facts of the bitcoin market”, Physica a-Statistical Mechanics and Its
Applications, 484, p. 82–90, 2017.
[14] D. Ron, A. Shamir, “Quantitative analysis of the full bitcoin transaction
graph”, International Conference on Financial Cryptography and Data
Security, pp. 6–24, 2013.
[15] V.D. Romanenko, Yu.L. Milyavsky. “Forecasting of maximal condi-
tional dispersions for multidimensional processes with mulirate dis-
cretization on the basis of adaptive GARCH models”, System Research
and Information Technologies, 2009, No. 4, pp. 92–108. (In Russian)
[16] B. Polyak, R. Shcherbakov. Robust Stability and Control. Moscow,
Nauka, 2002, 303 p. (In Russian)
[17] S. Nasin. B. Polyak, M. Topunov. “Suppression of constrained external
perturbations using the method of invariant ellipsoids”, Automatics and
telemechanics, 2007, No. 3, pp. 106–125.
444 Automated Control Problem for Dynamic Processes Applied to Cryptocurrency

[18] V.D. Romanenko, Yu.L. Milyavsky. “ Control automation of impulse

processes in cognitive maps with constrained disturbance suppression
based on invariant ellipsoids method”, System Research and Informa-
tional Technologies, 2017, no. 2, pp. 29–39. (In Russian)
[19] B.T. Polyak. “Convexity of quadratic transformations and its use in
control and optimization”, Journal of Optimization Theory and Appli-
cations, 1998, vol. 99, pp. 553–583.
[20] W.W. Hager. “Updating the inverse of a matrix”, SIAM Review, 1989,
vol. 31(2), pp. 221–239.
[21] V.D. Romanenko, Yu.L. Milyavsky. “Control automation method in
cognitive maps based on the synthesis of increments of weighting
coefficients and nodes coordinates”, System Research and Information
Technologies, 2019, no. 3, pp. 89–99. (In Russian)
[22] R. Isermann. Digital Control Systems: Springer-Verlag, Berlin, Heidel-
berg, 1981.
[23] K.J. Astrom, B. Wittenmark. Computer controlled systems, Prentice-
Hall, Inc., Englewood Cliffs, 1984.
[24] L. Ljung, “Systems identification. Theory for the user”, Prentice-Hall,
1986.
[25] D.W. Clarke, P.J. Gawthrop. “Self tuning control”, Proceedings of the
IEE, vol. 126, 1979, pp. 633–640.
[26] V. Gubarev. Modeling and Identification of Complex Systems, Kyiv,
Naukova Dumka, 2019, 247p. (In Ukrainian)
 Index

A Differential-difference games 113,

Algorithmic procedure 54, 289, 303, 138, 140
322 Discrete approximations 31, 35, 40,
Automatic control systems 32, 215, 48, 55, 57
294, 379, 414 Discrete time 40, 61, 63, 256, 419
Avoiding 5, 51, 97, 104, 263, 400 Dynamic games 114, 401, 415

B E
Bounded errors 167 Equilibrium problem 145, 148, 159,
Boundedness 63, 64, 67, 73, 76, 84, 163
91 Estimation 62, 88, 90, 154, 264, 370,
Braking distance 400 373, 407, 413
Estimation algorithm 83, 168
Euler–Lagrange and Hamiltonian
C inclusions 31
Complex discrete system 167 Evasion 4, 16, 104, 142
Conflict-controlled processes 3, 114, Extremal aiming rule 3
401
Constrained optimization 32, 380
Cryptocurrency 417, 418, 420, 421, F
423, 433 Feedback control system 65, 66, 89
Cyber-physical system 328, 329, Fuzzy controller 216, 218, 222, 264,
332, 333, 336 281
Fuzzy system 200, 202, 203, 224,
284
D
Data mining 236, 237, 380
Design space 265, 379, 380, 384, 394 G
Design and structural optimization Game theory 28, 128, 146
197, 204, 214, 228, 229 Games of approach 131, 138, 401,
Differential game 104, 113, 132, 142 415

445
446 Index

Generalized differentiation 31, 32, P

33, 35, 36, 41, 55 Petunin ellipsoids 379, 381
Group pursuit 4, 9, 17, 116, 142 Pontryagin’s maximum principle 31
Guaranteed result 4, 114, 131 Pontryagin’s first direct method 7, 8,
127, 139
Prediction set 379, 381, 389, 396
H Pseudoinverse model-based control
Hysteresis 31, 33, 39, 51, 58 62

I Q
Identification 77, 81, 99, 167, 169, Quadrotor UAV 289, 293, 294, 311,
257 312, 313
Information technology 197, 199, Quadrotor unmanned aerial vehicle
204, 207, 209, 210, 212, 229 197, 199, 214, 229
Integral inequality 100, 101, 105
Invariant ellipsoid method 417
R
Real-time algorithm 399, 415
L Regularization 145, 149, 167, 175,
Linear regression 167, 171 177
Linear matrix inequalities 292, 427 Regularized dimension 167
Linguistic terms 197, 198, 201, 204, Resolving function 7, 118, 121, 125,
210, 212, 219, 222, 229 135
Logistics 198, 328, 335, 346, 348 Rejected takeoff 399, 400, 404, 409,
412, 413
Riccati equation 341, 344, 399, 401,
M 413
Marine moving object 291, 292 Robust adaptive controller 62
Minimum trace ellipsoid 380 Robust-optimal control 289, 294,
Minimum volume ellipsoid 380 308, 309
Modal control 291, 418, 419, 423 Route planning 327

O
Optimal control 31–36, 42, 147, 289, S
294, 303, 309 Scheme of iterative regularization
Optimality conditions 31, 35, 43, 46, 145, 146, 159
294, 317, Set-valued mapping and its
Outlier detection 380 selections 3
 Index 447

State constraints 4, 17, 19, 20, Two-level problem 146, 147, 148,
48 116 159
State controller 423, 432 Two-stage proximal method 145
Stroboscopic strategy 6, 117
Strong convergence 41, 145, 160 U
Successive pursuit 3, 4, 6, 22, 25, 116 Uncertainty 4, 61, 66, 89, 114, 265,
SVD 174, 176, 177, 180 305, 422
Sweeping processes 31, 33, 38, 48,
59
V
Variational analysis 31, 33, 36, 43,
T 48, 55
Trajectory 4, 8, 25, 104, 182, 218, Variational inequality problem 145,
242, 301, 303 146, 158, 159
 About the Editors

Yuriy P. Kondratenko, Doctor of Science (habil.), Professor, Honour Inven-

tor of Ukraine (2008), Corr. Academician of Royal European Academy of
Doctors – Barcelona 1914 (2000), Head of Intelligent Information Systems
Department at Petro Mohyla Black Sea National University, Ukraine. Profes-
sor Kondratenko received a Ph.D. (1983) and a D.Sci. (1994) in Computer
and Control Systems from Odessa National Polytechnic University. He has
received several international grants and scholarships for conducting research
at the Institute of Automation of Chongqing University, P.R. China (1988–
1989), Ruhr-University Bochum, Germany (2000, 2010), Nazareth College
and Cleveland State University, USA (2003). In 2015 he received a Fulbright
grant for conducting research during 9 months at the Department of Electrical
Engineering and Computer Science, Cleveland State University, USA.
He is an author of more than 700 publications, including 140 patents and
14 books published by Springer, River Publishers, etc. He is a member of the
Scientific Committee of the National Council of Ukraine on Development
of Science and Technology, National Committee of Ukrainian Association
on Automatic Control, as well as visiting lecturer at the universities in
Rochester, Cleveland, Kassel, Vladivostok and Warsaw. He is a member of
Editorial Boards of such journals as Journal of Automation and Information
Sciences, International Journal of Computing, Eastern European Journal of
Enterprise Technologies, International Research and Review: Journal of Phi
Beta Delta, Quantitative Methods in Economics and others. His research
interests include intelligent decision support systems, automation, sensors
and control systems, fuzzy logic, soft computing, modeling and simulation,
and robotics.
Vsevolod M. Kuntsevich, Ph.D. (1959), D.Sci. (1965), Professor (1967),
Academician of the National Academy of Sciences of Ukraine (1992), Hon-
orary Director of Space Research Institute of NASU, Kyiv, Ukraine. Honored
Figure of Science and Technology of Ukraine (1999), laureate of the State
Prize of the Ukrainian SSR (1978, 1991) and Ukraine (2000) in the field
of science and technology, S. Lebedev Award (1987), V. Glushkov Award

449
450 About the Editors

(1995), V. Mikhalevich Award (2003). Professor Kuntsevich graduated from

Kyiv Polytechnic Institute (1952) and has worked at the Institute of Electrical
Engineering (1958–1963), Institute of Cybernetics (1963–1996), and Space
Research Institute (from 1996). He is part of the editorial staff of several
journals including Problemy Upravleniya I Informatiki and Cybernetics and
Systems Analysis. He is the founder of the national school in the field of
discrete control systems and he made a significant contribution to the devel-
opment of the modern theory of adaptive and robust control under uncertainty.
He is the author of 8 books and over 250 articles and during 25 years was
a Chairman of the National Committee of the Ukrainian Association on
Automatic Control (1993-2018).
Arkadii A. Chikrii, Ph.D. (1972), D.Sci. (1979), Professor (1989), Aca-
demician of the National Academy of Sciences of Ukraine (2018), Head
of the department “Optimization of Controlled Processes” of the Institute of
Cybernetics, NAS of Ukraine. Laureate of the State Prize of Ukraine in the
field of science and technology (1999), Glushkov Award (2003), Laureate of
the State Prize of Ukraine in the field of education (2018), Mikhalevich Award
(2021). After graduation from Ivan Franko Lviv University (1968), Professor
Chikrii has been working at the Institute of Cybernetics, Professor of Taras
Shevchenko Kyiv National University, Professor of Ihor Sikorski National
Technical University, and Professor of Yurij Fedkovich Chernivtsi National
University. He is the editor-in-chief of the journal Problemy Upravleniya I
Informatiki (since 2020). He is a specialist in the field of applied nonlinear
analysis, theory of extremal problems, mathematical theory of control, theory
of dynamic games, theory of search for moving objects, and computer tech-
nologies for analysis of conflict situations. He is a disciple and follower of
L.S. Pontryagin, N.N. Krasovskii and B.N. Pshenichnyi and is the author of
6 books and over 550 articles. He is a Chairman of the National Committee
of the Ukrainian Association on Automatic Control (since 2019).
Vyacheslav F. Gubarev, Doctor of Science (1992), Professor, Corresponding
Member of the National Academy of Science of Ukraine (2006), Professor
of Mathematics Methods of System Analysis Department, Kyiv National
Technical University, Ukraine, Head of Control Department, Space Research
Institute, National Academy of Science, Ukraine. Professor Gubarev received
a Ph.D. (1971) and a D.Sci. (1992) in System Analysis and Automatic
Control from the Institute of Cybernetics of the National Academy of Sci-
ence, Ukraine. He has taken part in several international grants with the
Russia Academy of Science, Moscow University, and others. His research
 About the Editors 451

interests include mathematical modeling of complex systems, automatic con-

trol, estimation and identification, ill-posed mathematical problems, dynamic
and control under uncertainty, and spacecraft control systems. He is part
of the editorial staff of several journals including Problemy Upravleniya I
Informatiki and Cybernetics and Systems Analysis. He is vicechairman of
the Ukrainian Association of Automatic Control, which is an NMO of IFAC.