Experience Seamless Full Ebook Downloads for Every Genre at ebookmeta.

com

AI Time Series Control System Modelling Chuzo

Ninagawa

https://ebookmeta.com/product/ai-time-series-control-system-
modelling-chuzo-ninagawa/

OR CLICK BUTTON

DOWNLOAD NOW

Explore and download more ebook at https://ebookmeta.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Advances in Power System Modelling, Control and Stability

Analysis 2nd Edition Federico Milano

https://ebookmeta.com/product/advances-in-power-system-modelling-
control-and-stability-analysis-2nd-edition-federico-milano/

ebookmeta.com

Modelling Trends and Cycles in Economic Time Series 2nd

Edition Terence C. Mills

https://ebookmeta.com/product/modelling-trends-and-cycles-in-economic-
time-series-2nd-edition-terence-c-mills/

ebookmeta.com

Mathematical Modelling Nonlinear Control and Performance

Evaluation of a Ground Based Mobile Air Defence System
Mechanisms and Machine Science 76 Constantinos Frangos
https://ebookmeta.com/product/mathematical-modelling-nonlinear-
control-and-performance-evaluation-of-a-ground-based-mobile-air-
defence-system-mechanisms-and-machine-science-76-constantinos-frangos/
ebookmeta.com

General, Organic, and Biological Chemistry: Structures of

Life 6th Edition Karen Timberlake

https://ebookmeta.com/product/general-organic-and-biological-
chemistry-structures-of-life-6th-edition-karen-timberlake/

ebookmeta.com
Demon Slayer Kimetsu No Yaiba Volume 14 Chapters 121 124
Koyoharu Got■ge

https://ebookmeta.com/product/demon-slayer-kimetsu-no-yaiba-
volume-14-chapters-121-124-koyoharu-gotoge/

ebookmeta.com

The Definitive Guide to Masonite: Building Web

Applications with Python 1st Edition Christopher Pitt

https://ebookmeta.com/product/the-definitive-guide-to-masonite-
building-web-applications-with-python-1st-edition-christopher-pitt/

ebookmeta.com

Textbook of Hospital Psychiatry 2nd Edition Harsh K.

Trivedi

https://ebookmeta.com/product/textbook-of-hospital-psychiatry-2nd-
edition-harsh-k-trivedi/

ebookmeta.com

Heartless (Hunter Moon Academy #3) 1st Edition Simone

Leigh Martin

https://ebookmeta.com/product/heartless-hunter-moon-academy-3-1st-
edition-simone-leigh-martin/

ebookmeta.com

Varieties of Precarity Melting Labour and the Failure to

Protect Workers in the Korean Welfare State 1st Edition
Seung Yoon Lee Sophia
https://ebookmeta.com/product/varieties-of-precarity-melting-labour-
and-the-failure-to-protect-workers-in-the-korean-welfare-state-1st-
edition-seung-yoon-lee-sophia/
ebookmeta.com
Mitochondrial Medicine: Volume 2: Assessing Mitochondria
Volkmar Weissig

https://ebookmeta.com/product/mitochondrial-medicine-
volume-2-assessing-mitochondria-volkmar-weissig/

ebookmeta.com
Chuzo Ninagawa

AI Time
Series Control
System
Modelling
AI Time Series Control System Modelling
Chuzo Ninagawa

AI Time Series Control

System Modelling
Chuzo Ninagawa
Smart Grid Power Control Engineering
Joint Research Laboratory
Gifu University
Gifu, Japan

ISBN 978-981-19-4593-9 ISBN 978-981-19-4594-6 (eBook)

https://doi.org/10.1007/978-981-19-4594-6

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Singapore Pte Ltd. 2023
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface

The Internet of Things (IoT) and artificial intelligence (AI) are without a doubt the
most important technology topics of the near future. As the world undergoes Digital
Transformation (DX), the Internet data collection through IoT is becoming the norm,
and an era of massive time series data accumulation is about to begin. Furthermore,
modeling technology will no longer be able to keep up with manually in extracting
relevant information from the large ocean of time series data. AI modeling will be
an inevitable core technology of the DX era.
Most image recognition, which is a representative of AI technology, can be said
to be static modeling that does not depend on past history, but the time series data
accumulated by DX can be said to be dynamic modeling in which the appearance of
values changes with past history. Long short-term memory (LSTM), a neural network
specialized for time series data, has been attracting attention as a neural network that
is good at prediction depending on the history of time series data, and its tools are
now readily available.
Since time series AI modeling is a sophisticated predictor that deals with history,
it is difficult for beginners to understand the learning theory and effective training
data collection techniques. For example, even if you study the theory and collect time
series data to test it in your own work, you will face problems such as obtaining only
biased training data or not being able to determine realistic convergence conditions.
However, there seem to be two extremes in the world: theoretical books with
only mathematical formulas, or how-to books on tools. For graduate students in
university laboratories and front-line engineers in the industrial world, there is a
need for specialized books that serve as a “bridge” by describing the development
of practical models based on theory.
In each chapter of this book, a structure is adopted that has never been seen before:
a section that presents the basic theory in mathematical form, followed by a section
that presents practical applications of the theory. In other words, the emphasis is on
showing concrete examples of the application of the basic algorithms in the field of
system control immediately after understanding them mathematically. By doing so,
the author aimed to take a different approach from mathematical books that develop
theories in an abstract manner by deriving pure mathematical formulas and from

v
vi Preface

how-to books that only describe how to input and output data to off-the-shelf tools
without describing theories.
In general, examples of AI machine learning may appear to be handled well,
but there may be cases where the reliability is not guaranteed. Therefore, all the
examples in this book were selected from peer-reviewed papers by IEEJ, IEEE, and
other experts to ensure reliability.
The structure of this book is as follows.
Chapter 1 begins with the definition of “time series” in system control and
describes the position of control design modeling from time series data of physical
quantities of interest.
Chapter 2 presents the basic theory of linear multiple regression modeling and
autoregressive (AR) modeling as the most fundamental methods in time series data
modeling and their practical applications.
Chapter 3 describes the basic theory of neural network modeling of dynamic
characteristics of control targets as a representative of AI machine learning modeling
with time series data and its application to modeling of step response characteristics.
Chapter 4 presents the theory of long short-term memory (LSTM) neural
networks, which have attracted attention in recent years as a method for modeling
control targets whose subsequent behavior differs depending on the time series
history, and an example of a prediction model for sudden events in control as an
application of the theory.
Chapter 5 describes the theory and examples of heuristic optimal search control
as optimal control using the above time series AI model.
Chapter 6 describes a practical method for collecting time series data in the field
of system control design, including a method for correcting collected data bias and
a method for estimating training data from normal operation data.
Chapter 7 describes a practical method for implementing the methods described
in the above chapters: a time series data collection platform, a method for extracting
zones of interest from field-collected data, and self-developed machine learning
software.
I would like to express my gratitude to many people for their help in compiling
this book. Morio Takahama, former professor at Nagoya University, as an expert in
control engineering, and Satoru Hayamizu, former professor at Gifu University, as
an expert in AI, provided valuable comments on the manuscript. Former Assistant
Professor Shun Matsukawa of Ninagawa Laboratory, Gifu University, and current
Lecturer at Hokkaido University of Science, helped to confirm the mathematical
expressions. Of course, I am grateful to Assistant Professor Yoshifumi Aoki, Ph.D.
student Asif Iqbal, and other members of the Ninagawa Laboratory at Gifu University
for their research. I would like to express my gratitude to all of them. Finally, I would
like to thank my wife for allowing me to write this book at home for a long time.

Gifu, Japan Chuzo Ninagawa

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Time Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1 What is “Time Series” Dealt in This Book? . . . . . . . . . . . . . . 1
1.1.2 Time Series for Statistical Control . . . . . . . . . . . . . . . . . . . . . . 2
1.1.3 Dissemination of Time Series Data for Control . . . . . . . . . . . 3
1.2 Time Series and Control Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2.1 Control Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2.2 Control Model Building Methods . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Control Time Series and AI Methods . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3.1 Control Model by Time Series Analysis . . . . . . . . . . . . . . . . . 6
1.3.2 Control and AI Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2 Linear Time Series Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1 Linear Regression Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1.1 One-Dimensional Linear Regression Model . . . . . . . . . . . . . . 9
2.1.2 Multi-Dimensional Linear Regression Model . . . . . . . . . . . . 12
2.2 Fundamentals of AR Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2.1 Overview of the AR Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2.2 Yule-Walker Method (One Variable) . . . . . . . . . . . . . . . . . . . . 16
2.2.3 Yule-Walker Method (Multivariate) . . . . . . . . . . . . . . . . . . . . . 18
2.3 Practical Example 1: Multiple Regression Model with Stable
Interval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3.1 Air-Conditioning Stable Power Model . . . . . . . . . . . . . . . . . . 21
2.3.2 Selection of Explanatory Variables . . . . . . . . . . . . . . . . . . . . . 23
2.3.3 Linear Multiple Regression Analysis . . . . . . . . . . . . . . . . . . . . 28
2.3.4 Model Evaluation and Validation . . . . . . . . . . . . . . . . . . . . . . . 29

vii
viii Contents

2.4 Practical Example 2: Step Response AR Model . . . . . . . . . . . . . . . . . 31

2.4.1 Limited Control of Building Air-Conditioning Power . . . . . . 31
2.4.2 Fitting the AR Mathematical Model . . . . . . . . . . . . . . . . . . . . 33
2.4.3 Model Identification from Measured Data . . . . . . . . . . . . . . . 35
2.4.4 AR Model Identification Results . . . . . . . . . . . . . . . . . . . . . . . 37
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3 Deep Learning AI Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.1 Fundamentals of Deep Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.1.1 Fundamentals of Neural Networks . . . . . . . . . . . . . . . . . . . . . . 41
3.1.2 Principles of Deep Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.1.3 Stacked Denoising Autoencoder Method . . . . . . . . . . . . . . . . 45
3.2 Time Series Data Deep Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.2.1 Time Series Parallel Input Neural Network . . . . . . . . . . . . . . 47
3.2.2 Number of Layers for Time Series Deep Learning . . . . . . . . 48
3.2.3 Hyperparameters for Time Series Deep Learning . . . . . . . . . 51
3.3 Practical Example 3: Step Response AR Neural Network . . . . . . . . . 54
3.3.1 Step Response AR Neural Network . . . . . . . . . . . . . . . . . . . . . 54
3.3.2 Training a Step Response Time Series Model . . . . . . . . . . . . 56
3.3.3 Evaluation of Step Response Time Series Models . . . . . . . . . 59
3.4 Practical Example 4: Deep Learning in Practice—Sudden
Event Prediction Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.4.1 Examples of Sudden Events . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.4.2 Sudden Event Prediction Neural Network Model . . . . . . . . . 61
3.4.3 Training a Neural Network Model for Sudden Event
Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4 LSTM AI Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.1 Fundamentals of LSTM Neural Networks . . . . . . . . . . . . . . . . . . . . . . 67
4.1.1 What is LSTM Neural Network? . . . . . . . . . . . . . . . . . . . . . . . 67
4.1.2 LSTM Forward Propagation Calculation . . . . . . . . . . . . . . . . 70
4.1.3 LSTM Back Propagation Calculation . . . . . . . . . . . . . . . . . . . 71
4.2 Performance Evaluation Methods for LSTM Time Series
Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.2.1 LSTM Model of Rare and Unexpected Events . . . . . . . . . . . . 73
4.2.2 Predictive Performance Evaluation Method . . . . . . . . . . . . . . 76
4.2.3 Results of Predictive Performance Evaluation . . . . . . . . . . . . 76
4.3 Practical Example 5: Electricity Wholesale Market LSTM
Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.3.1 Prediction of Wholesale Electricity Prices . . . . . . . . . . . . . . . 79
4.3.2 Electricity Wholesale Price LSTM Forecasting Model . . . . . 81
4.3.3 Evaluation of Wholesale Electricity Price LSTM
Forecasting Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Contents ix

4.4 Practical Example 6: LSTM Model for Prediction

of Disturbance Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.4.1 Example of a Time Series Unexpected Event . . . . . . . . . . . . . 84
4.4.2 Facility Maintenance Operation as a Disturbance
for RTP Adaptive Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.4.3 Disturbance Prediction LSTM Model for RTP
Adaptive Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.4.4 Evaluation of Disturbance Predictive LSTM Model
for RTP Adaptive Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5 Optimal Control Using Time Series AI Models . . . . . . . . . . . . . . . . . . . . 91
5.1 Fundamentals of Optimal Search and Control . . . . . . . . . . . . . . . . . . . 91
5.1.1 SA Optimal Search Method . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.1.2 Principle of Simulated Annealing (SA) Optimal
Search Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.1.3 Example of Evaluation Function for SA Optimal
Search Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
5.2 State Explosion and Parallel Search . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.2.1 Large-Scale Control Target State Space . . . . . . . . . . . . . . . . . 97
5.2.2 Parallel SA Search Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.2.3 Trials of Large-Scale Parallel Search . . . . . . . . . . . . . . . . . . . . 100
5.3 Practical Example 7: Electricity Price Optimal Search Control . . . . 105
5.3.1 Real-Time Electricity Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.3.2 Optimal Control of Air-Conditioning Power Rates . . . . . . . . 107
5.3.3 Actual Equipment Tests of Optimal Search Control . . . . . . . 110
5.4 Practical Example 8: Practical Cessation of Large-Scale
Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
5.4.1 Practicality of Optimal Search Control . . . . . . . . . . . . . . . . . . 114
5.4.2 Censoring of Large-Scale Search . . . . . . . . . . . . . . . . . . . . . . . 115
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
6 Reality of Time Series Learning Data Collection . . . . . . . . . . . . . . . . . . 121
6.1 Practical Example 9: Generating Training Data
with Pseudo-step Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
6.1.1 Step Response Training Data . . . . . . . . . . . . . . . . . . . . . . . . . . 121
6.1.2 Break-Point Step Response Extraction Method . . . . . . . . . . . 123
6.1.3 Example of Break-Point Method Training Data
Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
6.2 Practical Example 10: Artificial Augmentation of Training
Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
6.2.1 Reality of Training Data Collection . . . . . . . . . . . . . . . . . . . . . 130
6.2.2 Artificial Augmentation of Training Data . . . . . . . . . . . . . . . . 131
6.2.3 Practice of Artificial Training Data Augmentation . . . . . . . . 133
x Contents

6.3 Example 11: Generating Training Data with Emulators . . . . . . . . . . 136

6.3.1 Baseline and Reproducibility . . . . . . . . . . . . . . . . . . . . . . . . . . 136
6.3.2 Baseline Emulator Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
6.3.3 Baseline Estimation Model Evaluation . . . . . . . . . . . . . . . . . . 143
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
7 Practical Work on Time Series AI Modeling . . . . . . . . . . . . . . . . . . . . . . 147
7.1 IoT Time Series Data Collection Methods . . . . . . . . . . . . . . . . . . . . . . 147
7.1.1 IEEE1888 Standard for Time Series Data Collection . . . . . . 147
7.1.2 IEEE1888 Time Series Data Transmission Method . . . . . . . . 149
7.1.3 IEEE1888 Standard IoT Communication Software
Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
7.2 Zone Selection for Time Series AI Training Data . . . . . . . . . . . . . . . . 154
7.2.1 Reality of Time Series Training Data Collection . . . . . . . . . . 154
7.2.2 Zone Selection of Time Series Training Data . . . . . . . . . . . . . 156
7.2.3 Practical Methods with Time Series Data Selection . . . . . . . 159
7.3 Self-developed Software for Time Series AI Modeling . . . . . . . . . . . 161
7.3.1 Off-the-Shelf Training Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
7.3.2 Self-developing Machine Learning Software . . . . . . . . . . . . . 162
7.3.3 Visualization with Self-developed Machine Learning
Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
8 Example Source Code of MLP Deep Learning Algorithm . . . . . . . . . . 169
8.1 Execution Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
8.2 Example of MLP Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
9 Example Source Code of LSTM Neural Network Learning
Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
9.1 Execution Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
9.2 Baseline Estimation LSTM Time Series Learning Code
Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
About the Author

Prof. Chuzo Ninagawa is CEO of N Laboratory, Inc., and Professor of Smart Grid
Power Control Engineering Joint Research Laboratory¸ Gifu University, Gifu, Japan.
He has been Executive Chief Engineer of Mitsubishi Heavy Industries, Ltd., which is
one of the largest hi-tech manufacturers in Japan. His research interests span various
topics of smart grid, with special focus on virtual power plant (VPP) with a large-
scale aggregation of fast automated demand responses. He has published over 110
academic papers and five advanced research books.

xi
Chapter 1
Introduction

Abstract This chapter begins with the definition of “time series” in system control
and describes the position of control design modeling from time series data of phys-
ical quantities of interest. Statistical modeling approach is introduced as a black
box modeling on the control system target. Then, the concept of machine learning
modeling approach will be introduced.

Keywords Time series data · White box model · Black box model ·
Autoregressive model · Machine learning

1.1 Time Series

1.1.1 What is “Time Series” Dealt in This Book?

This book describes the practical application of artificial intelligence (AI) methods
using time series data in system control. It consistently discusses the application
of machine learning to the analysis and modeling of time series data of physical
quantities to be controlled in the field of system control.
In the field of system control, a time series is a sequence of physical quantity data
arranged in time order at fixed time intervals. Although it is not absolutely necessary
to have a fixed time interval, virtually any data sequence with a fixed time interval
is called a time series from the standpoint of mathematical handling and practical
data organization. Depending on the time scale of the physical quantity of interest,
the interval may be one microsecond or one hour. In any case, this book implicitly
refers to them as fixed time interval time series.
In recent years, the term “time series” has often been used in the field of informa-
tion science, especially in the field of AI, to refer to the order of appearance of words
and scenes in language processing and video processing. In contrast to control, the
term “time series” is used in situations such as word order and image transition.
These are different from time series in system control and are outside the scope of
this book.
The purpose of this book in discussing time series in system control is to predict
and manipulate the target physical quantity based on its history. A dynamic system
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 1
C. Ninagawa, AI Time Series Control System Modelling,
https://doi.org/10.1007/978-981-19-4594-6_1
2 1 Introduction

is a system in which the subsequent change in the physical quantity to be controlled

depends on the time series of the state physical quantity that determines the state.
Dynamic systems exhibit physical phenomena in which the transient characteris-
tics are problematic as well as the steady-state characteristics. Transients are prob-
lematic because they are affected by the system state before and after the change
occurs. The control target of this book is a system in which it is essentially important
to handle the time history, or time series, of the physical quantities that determine
the response behavior of the system state.
In addition, when dealing with time series for control, it is interesting to consider
target systems in which each physical quantity that constitutes the time series data
has a stochastic value.
In real-world control, even in systems where there is no uncertainty in the target
itself, it is natural for the observed data of the controlled quantity to be noisy, and
the actual manipulated physical quantity will naturally contain errors relative to the
calculated value of the manipulated quantity. However, the subject of interest to
control engineers is not merely such uncertainty in observations and operations, but
dynamic systems dominated by intrinsic stochastic phenomena.

1.1.2 Time Series for Statistical Control

Naturally, if the time series of the past state history of a dynamic system is different,
the transient response output, or the change in the controlled quantity, will be different
even if the current input, or the amount of operation, is exactly the same. In addition,
even if the time series of the state history and the input at the present time are exactly
the same, there are dynamic systems with uncertain intrinsic behavior, i.e., statis-
tical elements, depending on the target system. This book concentrates on statistical
dynamic systems with inherent uncertainty.
The uncertain behavior of such statistical dynamic systems cannot be easily
taken into account by the deterministic control theory that has been widely used
so far. In recent years, research results in control theory have led to the develop-
ment of advanced mathematical methods that can deal with uncertain control targets.
However, it must be said that these methods are too difficult for engineers in the field
who are involved in the control of embedded systems.
Such inherently uncertain dynamic systems may include the following. An
example of a statistical dynamic system is the prediction of the path of a typhoon. The
purpose of these systems is not to manipulate the system itself, but to quantitatively
estimate the transient changes in the future based on the past history of state. Load
frequency control, which targets each power plant from the power system’s central
power supply command center, is an example of statistical dynamic system control.
In these systems, prediction alone is not enough to solve the problem; it is necessary
to feed back the observed values and perform control operations to output the oper-
ational quantities. However, the system frequency, which is the controlled quantity,
cannot be completely controlled due to the inherent uncertainty, or statistical factor,
1.1 Time Series 3

of the control target, and the rate of stay of the controlled quantity in the target region
is to be kept within a practical range.
This section will review the target physical quantities, which are the controlled
quantity, the operating quantity, and the state variable in the statistical dynamic
system treated in this book. The physical quantities here are the quantities that deter-
mine the behavior of the target system, specifically, the properties governed by the
physical laws of natural science, such as temperature, voltage, and velocity. These
elements are called “physical control elements” in this book.
Next, we consider factors that are not “physically controlled” but are caused by
social factors. In general society, the word “control” is used to describe managerial
behavior toward highly uncertain phenomena. The word “infection control” for a
novel coronavirus, for example, is not the same as the restrictive definition of control
engineering. The term “control” seems to be closer to “management,” but the reason
to use the term “control” is that it involves some kind of feedback and manipulation to
approach a desired state. Naturally, these inherently uncertain elements are separate
from the laws of physics, such as the psychology of human behavior. These elements
are called “uncertain control elements.”
The control of statistical dynamic systems in this book deals with objects that have
both of the above elements. In other words, it deals with a very interesting system
control that combines two elements, one dealing with industrial physical quantities
and the other dealing with social uncertainties.
Although there are many aspects of complex system control consisting of uncer-
tain social and industrial systems that do not conform to the mathematical precision
of automatic control engineering, the controller outputs an operational quantity based
on the observed physical quantity of the dynamic system, and an uncertain response
is returned. Uncertainty management control, which increases the probability that
the controlled quantity will be in the desired range, is much needed in the real world.
In general, it must be said that automatic control theory as an academic discipline
has been extremely difficult to handle dynamic systems with uncertain responses,
although it makes full use of advanced mathematical theory. It is felt that a practical
method for managing and controlling a control target with such uncertain behavior
has not been accessible.
The topic of statistical extraction of “uncertainty control features” from time series
data is addressed, which has been increasingly popular in recent years, as one of the
answers in this book.

1.1.3 Dissemination of Time Series Data for Control

Now, let us discuss the history of time series data. Before the spread of digital
computers as we know them today, data to be controlled in system control was
often treated as analog quantities. In measurement and recording devices called pen
recorders, physical quantities measured by sensors were converted directly into pen-
driven electrical signals and recorded as curves on paper fed from a roll. It can be
4 1 Introduction

said that it is a kind of continuous measurement and recording of the time history of
physical quantities.
Microprocessors, or microcomputers, were introduced to the world around 1970.
It quickly developed and spread to the stage where it was called a personal computer
around 1980. Since then, microprocessors and personal computers have been actively
used in the measurement and recording of physical quantities for system control.
These digital computers became the norm, and pen recorders became extinct. In the
field of system control, measurement and recording came to be handled as digital data
in the form of discrete numbers. As a natural progression, it became common to record
discrete numerical values at fixed time intervals, and time series data became the
basic method of recording the time history, time changes, and transient phenomena
of physical quantities to be controlled.
Since around 1980, with the spread of microcomputers, a large number of devices
and systems in society have come to be controlled by embedded systems. However, at
that time, the embedded control of various equipment systems was developed within
the equipment itself, and there were few systems that continuously communicated
operation data to the outside for a long period of time. From around 2000, remote
centralized control of equipment systems began to be used, and cases where data on
equipment behavior was accumulated as a time series gradually appeared.
In addition, the Internet of things (IoT) communication network has become
widespread since around 2010, and it has become possible to accumulate time series
data on the behavior of embedded control systems on the things side of the IoT easily
and at low cost. On the Internet side of the IoT, the cloud centralized monitoring and
control side constantly collects time series data from many things side.
In this way, we are living in an era in which time series data on the behavior of
a huge variety of devices is constantly being collected and stored in data centers
at a short-time granularity through the IoT, and it seems that time series data is
now regarded as a natural resource that is collected and stored even without saying
anything.
As well as, we are living in an era where it is normal to collect and accumulate
time series data on the behavior of control targets. It is becoming more and more
important for engineers in the field to effectively use this vast amount of accumulated
time series data for system control design, especially for modeling the control target.

1.2 Time Series and Control Models

1.2.1 Control Modeling

In the field of system control, a model is a representation of the dynamic charac-

teristics of a system in a specific form, focusing on the essential parts from the
perspective of control. This is too abstract, so let us consider it more concretely from
the viewpoint of system control.
Other documents randomly have
different content
 Bore it with a patient grace;
By no boastful pride elated,
When he conquered time and space.”
 INDEX.
A
Acoustic figures, 117
Alpine adventures of Professor Tyndall, 60
Alps, accidents on, 65
America, electrical discoveries in, 231;
first line of telegraph in, 273;
telegraph in (see Morse and Telegraph);
visit of Professor Tyndall to, 74
Ampère’s electrical discoveries, 91;
proposed telegraph, 134
Aqueous vapour and radiant heat, 44
Arago’s electrical discoveries, 91
Atlantic cable, 193, 276, 292
Automatic telegraph, Wheatstone’s, 199

B
Bain, Alexander, inventions claimed by, 160, 185, 305
Baltimore and Washington telegraph, 273
Batteries described—Volta’s, 88;
Grove’s, 89;
Daniell’s, 128
Beer disease, 53
Bible descriptions of nature, 6
Biographies, use of, xi., xiv.
Blackwall telegraph, 167, 173
Brewster, Sir David’s account of first telegraph, 150;
on vision, 210;
improvement of stereoscope, 212
Bridge, Wheatstone, 164
Bryant, W. Cullen, on Morse and his telegraph, 252, 316

C
Cables, earliest, 187, 269, 292
Calorescence, 47
Carlyle, Thomas, reminiscences of, 99
Celtic genius and science, 7
Channel cable, first, 187, 191
Charges for telegraphing, 181
China, telegraph to, 317
Clark, Latimer, on first English telegraph, 152;
on Wheatstone’s single-needle telegraph, 167;
on Wheatstone’s works, 229
Clock, Wheatstone’s electro-magnetic, 160
Clouds, experiments in producing, 49
Concertina, invention of, 120
Congress, American, and telegraph, 263, 270
Cooke, W. F., account of his first connection with telegraph, 150,
152;
dispute with Wheatstone about telegraph, 134, 138, 146;
efforts to extend telegraph, 173;
formation of Electric Telegraph Company, 183
Cruikshank, George, on first telegraph, 141
Cryptograph, invention of, 219
Crystals, formation of, 96;
magnetic properties of, 24

D
Daniell, Professor, on Wheatstone’s first telegraph, 149
Daniell’s constant battery, 128
Day, Professor J., electrical lectures, 234
Dial telegraphs, Wheatstone’s, 158, 196
Diamagnetism discovered, 23;
investigated by Tyndall and others, 24, 29, 33
Dynamic radiation of heat, 43
Dynamo machine, invention of, 206

E
Earth as return circuit, 171;
rotatory motion, 217
Earth’s magnetic force, 26
Electric currents, measurement of, 93, 163
Electric telegraph. See Telegraph.
Electric Telegraph Company, formation of, 183
Electrical biographies, use of, xi.
Electrical heat and light, 89
Electricians, distribution of, xii., 231
Electricity, production of, 22, 88, 91, 94, 232;
force of, 163;
velocity of, 93
Ellsworth, Miss, connection with Morse telegraph, 272, 277
Enchanted lyre, Wheatstone’s, 111
Evolution, early days of Darwinian theory of, 97
Exploder, Wheatstone’s, 186
Explosion of mines by electricity, 185

F
Faraday’s associations with Professor Tyndall, 26, 30, 102;
electrical and magnetic discoveries, 23;
lecture on scientific theories, 32;
on Wheatstone’s telegraph, 198
Forbes, Professor J. D., on glaciers, 37, 40
Frankland, Dr., associated with Professor Tyndall, 15;
glacier theory by, 45

G
Gale, Professor, assisted Morse with telegraph, 249
Gases, radiation and absorption of heat by, 42;
sounding power of, 58
Gauss and Weber’s telegraph, 136
Germ theory, 51, 98
German scientists, 21, 27
Germany, science in, between 1840 and 1850, 16;
student life in, 17;
telegraph in, 136
Glacier phenomena, 38

H
Harmonium, Wheatstone’s improvements in, 123
Heat, radiant, investigation of, 42, 58
House, R. E., printing telegraph by, 306

I
Induced electricity, discovery of, by Faraday, 22
Inventions, popular accounts of origin of, 167;
Morse’s definition of, 280;
public appreciation of, 281, 302
Irish scientists, 7

J
Jackson, Dr., disputes with Morse origin of telegraph, 244, 256

K
Kaleidophone, 117

L
Light, velocity of, 125
Lightning conductors, 131
Longitude determined by telegraph, 287

M
Magnetic attraction, 28
Magnetic exploder, 186
Magnetisation of light, 93
Magnetism and diamagnetism, 23, 29;
of the earth, 26
Magnetism and electricity, 22, 91
Magnetism, mechanical theory of, 93
Magneto-electric machine, Wheatstone’s, 159
Magnets, interaction of, 91;
lengthened by electricity, 92
Marburg, student life in, 17
Measurement of electric currents, Wheatstone’s plans for, 124, 163
Metals, new, discovered by electric spark analysis, 127
Microphone, first use of word, 119
Morse alphabet, uses of, 307
Morse, Professor S. F. B.:
artist, how he became an, 236;
success as, 243;
why he ceased to be an, 279
Atlantic cable, connection with, 276, 292
birth and education of, 233
Congress’s action towards, 263, 270
death of, 320
difficulties in constructing his telegraph, 246;
in introducing it, 268, 281
electrical studies, 234, 242, 244
first line of telegraph constructed by, 273
funeral of, 321
honours conferred on, 308, 311
Jackson, Dr., controversy with, 244, 256
law-suits to protect patent rights, 303
London visited by, 236, 295
patents, 259, 265;
defence of, 303
pictures painted by, 237
photography, early connection with, 266
proscribed German student’s case, 253
rewards of, 309, 313
statue of, 315
telegraph, distinguishing features of, 279;
first conception of, 244;
 first public description of, 260;
labours to improve, 247;
practical working of, 260;
public trial of, 268;
refusal of, by American Government, 283;
spread of, 284;
uses of, 286;
working of, 289
trial of first telegraph line, 277
submarine cable, first, 269, 276, 292

N
Needle telegraph, 143, 167
Niagara visited by Prof. Tyndall, 75

O
Ohm’s work and theory, 140
O’Shaughnessy, Dr., introduction of telegraph in India, 310

P
Palmerston, Lord, on telegraph, 194
Pasteur’s experiments with germs, 52
Photography, invention of, 266;
introduction of, 211, 267
Piz Morteratch, accident upon, 67
Polarised light, Wheatstone’s experiments, 222
Printing telegraph, Wheatstone’s, 161
Proscribed German student, Morse’s account of, 253
Pseudoscope, invention of, 216

Q
Queenswood College, 15

R
Railway mania of 1845, 13
Recording telegraph, Morse’s, 277, 290;
Wheatstone’s, 199
Relay, first accounts of, 141, 249
Resistance measurer, 163
Return circuit, 171
Revolution effected by electricity, ix.
Revolving mirror, uses of, 124
Rheostat, Wheatstone’s, 165
Ricardo, J. L., connection with telegraph, 183
Ronalds’s telegraph, 110
Rosa, Monte, ascent of, 61
Royal Institution, changes at, 84;
lectures by Tyndall at, 30, 38, 87
S
Scientific attainments, recognition of, in England, 35
Scientific discovery, the pursuit of, 79
Sea-water, varying tints of, 56
Semaphore telegraph, 180
Slaty cleavage, 36
Smoke respirator, invention of, 54
Sound, transmission of, 56;
Wheatstone on, 116
Sounder, the Morse, 291
Spectrum analysis of electric spark, 127
Standards, electrical, 164
Steinheil’s telegraph, 136
Stereoscope, invention of, 210;
improvement of, 212;
principle of, 215
Submarine cables, earliest experiments with, 187, 269, 276, 292

T
Tawell, murderer, apprehended by use of telegraph, 178
Telegraph, adoption of, by public, 173, 283
automatic telegraph of Wheatstone, 199
cables, earliest, 189, 269, 292;
illustration of working, 95
charges for, 181
dial, invented by Wheatstone, 158;
improvement of, 196
early forecasts of, 106;
early achievements of, 173, 277, 284
electro-magnetic, Morse’s, 248, 277, 290;
Wheatstone’s, 158
extension of, 173, 181, 284, 292
history of, 134, 144, 153, 173, 244, 260, 282, 292
idea and invention of, 105, 244
longitude ascertained by, 287
Morse’s recording, 244, 260, 280, 290
needle, 143, 167
origin of, 134, 138, 142, 150, 244, 292
pedigree of, 108
recording, 199, 246, 260, 268, 277, 290
relay, 141, 249
sounder, the Morse, 290
Wheatstone’s first needle, 143;
 dial, 158;
printing, 161;
recording automatic, 199
Telephone, first, 115
Thermo-electric pile, 129, 205
Thermometers, self-registering, 221
Tyndall, Professor J.:
ancestors of, 3
anecdotes of, 34, 93, 97
birth and education of, 4
daring experiment by, 47
description of, by George Ripley, 72
diamagnetism, explanation of, 24, 29
duty, sense of, 19
endowments for scientific purposes, 80
Faraday, associations with, 26, 30, 102
Germany, student life in, 17, 21
German scientific friends of, 21, 26
investigation of diamagnetism, 24, 29;
germs, 51, 98;
glacier phenomena, 38;
radiant heat, 42, 58;
sea-water tints, 56;
slaty cleavage, 36;
sound, 56
marriage of, 86
Ordnance Survey joined, 9
Pasteur, remarks on, 52
pecuniary assistance declined by, 20, 102
Presidential address to British Association, 81
Professor of Natural Philosophy, appointed, 31
radiant heat, on, 42, 58
railway surveying by, 12
reminiscences of Thomas Carlyle, 99
 Royal Institution, at, 30, 85
scientific adviser to Trinity House, 102
scientific examiner at Woolwich, 35
smoke respirator, invention of, 54
teaching at Queenswood College, 15;
elsewhere, 96
travels of, in the Alps, 60;
at Vesuvius, 70;
in America, 71, 74
vindication of scientific education, 35.
working habits, 12
youthful studies, 8, 10, 21

V
Velocity of electricity, 93, 124;
of light, 125
Vesuvius, visited in 1868, 70
Vision, Wheatstone’s elucidations of, 210
Voltaic battery described, 88;
discovered, 110

W
West, Benjamin, associated with Morse, 236
Wheatstone, Professor Charles:
birth of, 111
bridge, 164
cryptograph, 219
death and funeral, 228
deciphering secret document, 220
dispute with W. F. Cooke about telegraph, 134, 138, 146, 153
electricity, first studies in, 123
enchanted lyre of, 111
harmonium improvements, 123
honours conferred on, 166, 226
invention of chronoscope, 162;
concertina, 120;
cryptograph, 219;
dynamo, 206;
electric clock, 160;
enchanted lyre, 111;
kaleidophone, 117;
magnetic exploder, 186;
magneto-electric machine, 159;
polar clock, 223;
pseudoscope, 216;
stereoscope, 210;
telegraph, 134 (see Telegraph);
thermometers, 221
inventions, periodicity of, 223
investigation of algebra, 224;
Chladni figures, 117;
earth’s motion, 217;
mental philosophy, 117;
musical instruments, 120;
polarised light, 222;
sound, 116, 118;
submarine cables, 187;
submarine explosions, 185;
thermo-electric pile, 129, 205;
tone, 224;
vision, 210
investigations, latest and incomplete, 224
lightning conductors, opinions on, 131
magnetic exploder, 186
measurement of force of electric currents, 163
originality of his telegraph, 134, 138, 144
patents of, 142, 154, 160, 167, 196
peculiarities of, 225
Professor of Experimental Physics at King’s College, 123
 revolving mirror, 124
speaking machines, improvements in, 117
spectrum analysis of electric light, 127
submarine cables, early experiments with, 187
telegraph, diagram of first, 141;
history of, 144, 153;
origin of, 134, 138, 142, 153
telegraphic instruments, automatic, 199;
dial, 158, 196;
needle, 141, 145, 167;
printing, 161
thermo-electric pile, 129, 205

RICHARD CLAY AND SONS, LONDON AND BUNGAY.

New Series,
No. 8.
2, White Hart Street,
Paternoster Square,
London, E.C.

WHITTAKER AND CO.’S

NEW PUBLICATIONS.

LIVES OF THE ELECTRICIANS.

First Series.

Professors TYNDALL, WHEATSTONE, AND MORSE.

BY

WILLIAM T. JEANS.

Crown 8vo. 6s.

The first volume of a series of Lives of the Electricians. It will contain popular
biographies of Professors Tyndall, Wheatstone, and Morse, telling incidentally the
story of the progress of the electric telegraph from its origin in 1837 to the present
time. Next year will be the jubilee of the electric telegraph in England.

“THE SPECIALISTS’ SERIES.”

New Volume, Just Published:—

Heat.
ON THE CONVERSION OF HEAT INTO WORK. A Practical
Handbook on Heat-Engines. By William Anderson, M.Inst.C.E.
With 55 Illustrations. Crown 8vo, pp. viii-252, 6s. cloth.
The object of this work is to popularize the doctrine that, in heat-engines, the
work given out is due to the conversion of the molecular motion of heat into the
visible motion which it is desired to produce; and further to illustrate, by numerous
practical examples, the applicability of the doctrine of Sadi Carnot to defining the
limits within which improvement in the economical working of heat-engines is
possible.

THE TELEPHONE AND ITS PRACTICAL APPLICATIONS. By W. H.

Preece, F.R.S., and J. Maier, Ph.D. [Ready shortly.
 “THE SPECIALISTS’ SERIES.”

A New Series of Handbooks for Students and Practical Engineers.

Crown 8vo. cloth. Illustrated throughout with original
and practical Illustrations.

Now Ready.

Electric Lighting.
ARC AND GLOW LAMPS. A Practical Handbook on Electric
Lighting. By Julius Maier, Ph.D., Assoc. Soc. Tel. Eng., etc.
With 78 Illustrations. Crown 8vo, pp. viii-376, 7s. 6d. cloth.
The whole system of modern electric [Just Published.
illumination is dealt with in this volume. It gives a detailed
description of all the principal modern generators and lamps,
together with conductors and the other appliances required in
electric light installations.
It contains also a full account of the various Applications of
Electric Lighting up to recent date.
“The author has collected all the most recent available information concerning
the process of manufacture, life, &c., of arc and glow lamps in a very convenient
and readable form. Indeed, we do not know any work in which the subject is, on
the whole, so fully handled.”—The Engineer.

Electric Transmission of Energy.

ELECTRIC TRANSMISSION OF ENERGY, and its
Transformation, Subdivision, and Distribution. A Practical
Handbook by Gisbert Kapp, C.E., Associate Member of the
 Institution of Civil Engineers, &c. With 119 Illustrations.
Crown 8vo, pp. xi-331, 7s. 6d. cloth.
∵ It has been the aim of the author to present the scientific part of
the subject in as simple a form as possible, giving descriptions of
work actually carried out. He has endeavoured in this way to place
before the reader an unbiassed report on the present state of
electric transmission of energy.
“This is ‘a practical handbook’ par excellence—a book which will be read,
studied, and used not by electricians merely, but by most engineers. It contains a
vast amount of original matter, and it bears the signs of much patient thought
assisted by practical experience.—
“We cannot speak too highly of this admirable book, and we trust future
editions will follow in rapid succession.”—Electrical Review.
“A valuable work; written with regard to facts only.”—Electrician.

Gas Motors.
GAS ENGINES. Their Theory and Management. By William
Macgregor. With 7 Plates. Crown 8vo, pp. 245, 8s. 6d. cloth.
List of Contents.
Introductory—Direct Working Engines without Compression—Gas Engines
working with Compression—Compression Engines with Compressing Pump—
Theory of the Gas Engine—Relative Speed of Combustion in Gaseous Explosive
Mixtures—Witz’s Theoretical Cycles of Gas Engines—Some further Theoretical Data
—Clerk’s Theory of the Gas Engine—The Gas Engine Indicator-Diagram—Index.
“Mr. Macgregor has collected a considerable amount of information on his
subject of a kind which may prove valuable to many readers. All who desire to be
well informed in gas engines will be able to get what they want from these
pages.”—Engineering.
“From the Abbé Hautefeuille’s powder machine, invented in 1678, to the Maxim
Patent of 1883, is a long record of progress fully detailed in Mr. Macgregor’s useful
and interesting book.”—Saturday Review.
“This handbook may be safely recommended to all students who wish to
acquire a thorough practical knowledge of the mechanism and theoretical
principles of the gas engine.”—The Building News.
Electro Motors.
MAGNETO- AND DYNAMO-ELECTRIC MACHINES. With a
Description of Electric Accumulators. From the German of
Glaser de Cew. With 61 Illustrations. Crown 8vo, pp. xiii-301,
6s. cloth.
In successive chapters the author considers electrical units; the
historical development of magneto- and dynamo-electric generators;
machines generating alternating and direct currents; the particular
applicability of the various electric generators; automatic switches
and current regulation; electrical storage; the physical laws bearing
on the construction of electric generators; the construction of the
several parts of electric generators; the employment of these
machines in producing the electric light; and for various other
purposes.
“Almost all the best known machines are described and illustrated, with the
discussion of certain theoretical questions.”—Electrician.
“Will be welcomed by those who wish, without studying the matter for
professional purposes, to possess a scientific knowledge of modern electrical
machines.”—English Mechanic.
“Presents in condensed form an epitome of electrical progress up to recent
dates.”—Scientific American.

Ballooning.
BALLOONING: A Concise Sketch of its History and Principles.
From the best sources, Continental and English. By G. May.
With Illustrations. Crown 8vo, pp. vi-97, 2s. 6d. cloth.
∵ This deals, not with the possibilities of aëronautics on vague
assumption, but gives information from a practical view of what has
been done, showing the present position.
List of Contents.
 Introduction—First Practical Experiments—Resources and Incidents—The
Practical Application of Aëronautics—Military Applications of Ballooning—Steering
Power—Present State of Ballooning and Recent Proposals for Steering Balloons—
The Cost.
“Mr. May gives a clear idea of all the experiments and improvements in aëro-
navigation from its beginning, and the various useful purposes to which it has
been applied.”—Contemporary Review.
“It confines itself to the statement of facts, and should fulfil completely the
purpose for which it was written.”—The Graphic.

JUST PUBLISHED.

THEORY
OF

MAGNETIC MEASUREMENTS,
WITH AN APPENDIX
ON THE

METHOD OF LEAST SQUARES.

BY
FRANCIS E. NIPHER, A.M.,
Professor of Physics in Washington University, St. Louis, and President of
the St. Louis Academy of Science.

One Volume. Crown 8vo. 5s.

 Will be Published early in January.

One Volume, royal 4to, with 21 double and 30 single

Plates.

TECHNICAL SCHOOL AND

COLLEGE BUILDINGS:
BEING

a treatise on the design and construction of applied

science and art buildings, and their suitable
fittings and sanitation,

With a Chapter on Technical Education.

BY
EDWARD COOKWORTHY ROBINS, F.S.A.,

Fellow of the Royal Instit. of Brit. Architects, Member of the Institute of Surveyors,
Member of Council of the Sanitary Institute of Great Britain, &c. &c.,
Member of the Executive Committee of the City and Guilds
of London Institute for the Advancement of
Technical Education.

CONTENTS.
CHAP.
Introduction.
I. Our British and Foreign Technical Education.
 II. An Analysis of the Second Report of the Royal Commissioners
on Technical Education.
III. Buildings for Applied Science and Art Instruction. Illustrated
by Foreign examples.
IV. Buildings for Applied Science and Art Instruction. Illustrated
by English examples.
V. Fittings necessary for Applied Science Instruction.
VI. A Description of the various Drawings illustrating the Fittings
required for Science Buildings.
VII. Heating and Ventilation as required for such Buildings, and
Belgian mode of Calculation.
VIII. Special examples of Heating and Ventilation.
IX. Buildings for Secondary Education generally (with discussion
thereon).
X. Sanitary Science in its relation to Civil Architecture (with
discussion thereon).

LONDON:
WHITTAKER AND CO., 2, WHITE HART STREET,
PATERNOSTER SQUARE, E.C.
 The Cheapest and the Best Office Book, handsomely bound in fcap. 8vo.

Will be Published early in January.

DOD’S PEERAGE, BARONETAGE,

AND KNIGHTAGE
OF
GREAT BRITAIN AND IRELAND,
FOR 1887,
INCLUDING
ALL THE TITLED CLASSES.

Forty-Seventh Year.

This differs from all other Peerages in—I. Its low price. II. Its enlarged Contents. III. Its
facility of reference.

Cloth, gilt, 10s. 6d.

JUST PUBLISHED.
A New Edition containing all the recent Parliamentary
changes to August 20th.

DOD’S
PARLIAMENTARY COMPANION.

60th issue. 1886. Second Edition.

Neatly bound, gilt edges, 4s. 6d.

 SURTEES’ SOCIETY’S PUBLICATIONS.

New Volumes. Just Published.

YORKSHIRE DIARIES AND AUTOBIOGRAPHIES IN THE 17th
and 18th CENTURIES. With Portraits, 8vo, pp. ii-173. Cloth. 7s. 6d.
MEMORIALS OF THE CHURCH OF SS. PETER AND WILFRID,
RIPON. Vol. II. 8vo, pp. xii-398. Cloth. 15s.
 JUST PUBLISHED.
A Cheap Edition of Nimrod’s Celebrated Letters.
NIMROD’S REMARKS ON THE CONDITION
OF HUNTERS.
The Choice of Horses and their Management.

By C. TONGUE.
Fourth Edition, tastefully bound in cloth, 2s. 6d.

PRACTICAL

MERCANTILE CORRESPONDENCE.
A COLLECTION OF
COMMERCIAL LETTERS AND FORMS
WITH EXPLANATORY NOTES, INDICATING THE
CORRECT EQUIVALENTS IN FRENCH,
GERMAN, OR ENGLISH,
AND A
VOCABULARY OF COMMERCIAL TERMS.
EDITED BY
L. SIMON, CHR. VOGEL, Ph.D., H. P. SKELTON,
w. c. wrankmore, leland mason,

and others.

Now Ready, crown 8vo, cloth.

ENGLISH, with German Notes. 3s.

GERMAN, with English Notes. 3s.
ENGLISH, with French Notes. 4s. 6d.
FRENCH, with English Notes. 4s. 6d.
 TECHNOLOGICAL HANDBOOKS.

Edited by H. TRUEMAN WOOD,

B.B. Secretary to the Society of Arts.
A Series of Technical Manuals for the use of Workmen and others practically
interested in the Industrial Arts, and specially adapted for Candidates in the
Examinations of the City Guilds Institute.
Illustrated and uniformly printed in small post 8vo.
1. DYEING AND TISSUE-PRINTING. By William Crookes,
F.R.S., V.P.C.S. 5s.
2. GLASS MANUFACTURE. Introductory Essay by H. J.
Powell, B.A. (Whitefriars Glass Works); Crown and Sheet Glass,
by Henry Chance, M.A. (Chance Bros., Birmingham); Plate
Glass, by H. G. Harris, Assoc. Memb. Inst. C.E. 3s. 6d.
3. COTTON SPINNING: Its Development, Principles, and
Practice. By R. Marsden, Editor of the “Textile Manufacturer,”
with an Appendix on Steam Engines and Boilers. 6s. 6d.
4. COAL-TAR COLOURS, The Chemistry of. With special
reference to their application to Dyeing, &c. By Dr. R. Benedik,
Professor of Chemistry in the University of Vienna. Translated
from the German by E. Knecht, Ph.D., Head Master of the
Chemistry and Dyeing Department in the Technical College,
Bradford. 5s.

Ready, 2 vols., 8vo, pp. 711-970, cloth, £1 7s.

TECHNOLOGICAL DICTIONARY
of the
ENGLISH AND GERMAN LANGUAGES.
Containing Words and Phrases employed in Civil and Military
Engineering, Shipbuilding and Navigation, Railway Construction,
Mechanics, Chemistry, Chemical Technology, Industrial Arts,
Agriculture, Commerce, Mining, Physics, Meteorology, Metallurgy,
Mathematics, Astronomy, Mineralogy, Botany, &c.
In Co-operation with P. R. Bedson, O. Brandes, M. Brütt, Ch. A.
Burghardt, Th. Carnelly, J. J. Hummel, J. G. Lunge, J. Lüroth, G.
Schäffler, W. H. M. Ward, W. Carleton Williams.
Edited by GUSTAVUS EGER,

Professor of the Polytechnic School of Darmstadt, &c.

Vol. I.—English-German.
Vol. II.—German-English.

“A really valuable work, which treats the two languages well and exhaustively,
and, best of all, correctly. We can confidently recommend it to every one who has
to work in English and German technical terms.”—Engineering.
 Just published, Vol. I., English-Spanish, in super-royal 8vo, pp. 873,
bound in half-morocco, £1 16s.

TECHNOLOGICAL DICTIONARY.
ENGLISH-SPANISH AND SPANISH-ENGLISH.
Containing Terms employed in the Applied Sciences, Industrial
Arts, Mechanics, Fine Arts, Metallurgy, Machinery, Commerce,
Ship-building and Navigation, Civil and Military Engineering,
Agriculture, Railway Construction, Electro-technics, &c.
By NESTOR PONCE DE LEON.

Vol. I.—ENGLISH-SPANISH.
Vol. II.—SPANISH-ENGLISH. [In Preparation.

Post 8vo, cloth, pp. xii-203, price 5s.

A BIBLIOGRAPHY OF ELECTRICITY AND
MAGNETISM,
1860 to 1883.
With Special References to Electro-Technics.
Compiled by G. MAY.
With an Index by O. Salle, Ph.D.

Cloth, price 2s. 6d. each; the Two Parts in one Volume, 5s.
Part I. ENGLISH-GERMAN.
Part II. GERMAN-ENGLISH.
TECHNOLOGICAL DICTIONARY
 Of the Physical, Mechanical, and Chemical Sciences.
By F. J. Wershoren, D.Sc.
“It is worth having, and by its cheapness in the reach of everybody.”—Textile Colourist.

NEW EDITION.
FODEN’S MECHANICAL TABLES
For Smiths and Millwrights.
Fourth Edition, Crown 8vo, pp. 47, 1s. 6d. cloth.

LONDON:
WHITTAKER AND CO., 2, WHITE HART STREET,
PATERNOSTER SQUARE, E.C.