Scribd
Upload
4 views

session 10 -P5_Robust Control Design

The document outlines a course on designing robust controllers for microgrids, focusing on the Robust Servomechanism Problem (RSP) and its applications. It introduces systematic methods for control systems that can be utilized by engineers with basic knowledge, emphasizing the design of servo-compensators and stabilizing compensators. The course also covers Linear Quadratic Tracking (LQT) and provides examples relevant to microgrid control.

Uploaded by

Arash Torkaman
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
4 views

session 10 -P5_Robust Control Design

The document outlines a course on designing robust controllers for microgrids, focusing on the Robust Servomechanism Problem (RSP) and its applications. It introduces systematic methods for control systems that can be utilized by engineers with basic knowledge, emphasizing the design of servo-compensators and stabilizing compensators. The course also covers Linear Quadratic Tracking (LQT) and provides examples relevant to microgrid control.

Uploaded by

Arash Torkaman
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 19

Smart Microgrids–ELE6427

Design of Robust Controllers and Applications in Microgrids

1
2

Outline

 Introduction
 Servomechanism Problem
 Servocompensator and State Feedback
 Linear Quadratic Tracker (LQT)
 Examples in Microgrid Control
3

Introduction
Most control systems are described by Multi-input Multi-
output (MIMO) models. Although several control methods
have been developed for such systems, there is no
systematic and easy to apply method that can be used by
an engineer with basic knowledge of control theory.

The existing methods, e.g., 𝐻𝐻∞ , 𝜇𝜇-synthesis, Lyapunov


design, … require deep knowledge of advanced
mathematics and control theory.

In this course, we try to introduce a systematic method


which can be applied to most Microgrid systems, and can
be employed by users with basic control knowledge!
4

Introduction
The proposed method is a special case of Robust
Servomechanism Problem (RSP) developed by Prof. E. J.
Davison (Emeritus Prof. of University of Toronto) in 70’s and
80’s. The interested readers are invited to study the
following papers:

[1] E. Davison and I. Ferguson, "The design of controllers


for the multivariable robust servomechanism problem using
parameter optimization methods," in IEEE Transactions on
Automatic Control, vol. 26, no. 1, pp. 93-110, Feb 1981.
[2] D. Davison, E. Davison, “Optimal Transient Response
Shaping of the Servomechanism Problem,” Journal of
Optimization Theory and Applications (2002) 115: 491.
5

Robust Servomechanism Problem


Structure of robust servomechanism control (RSC):
1- Servo-compensator: ensures asymptotic tracking of
reference signals, and rejection of disturbances.
2- Stabilizing compensator: guarantees closed-loop stability
and other desired performance characteristics.
Stabilizing Compensator w

yref + e Servo- xc
Kc
- u
Plant
y
compensator
- -
xp
Kp
6

RSP Formulation
Plant is LTI and its mathematical model is:

where: and
are real matrices of appropriate size.

Assumption I: Plant is stabilizable and detectable.


The reference signals and disturbances are produced
by the following dynamical systems.
7

Servo-compensator Structure

Without loss of generality, assume that (𝐶𝐶𝑟𝑟𝑟𝑟𝑟𝑟 , 𝐴𝐴𝑟𝑟𝑟𝑟𝑟𝑟 )


and (𝐶𝐶𝑤𝑤 , 𝐴𝐴𝑤𝑤 ) are observable.
Let the minimal polynomials of matrices 𝐴𝐴𝑟𝑟𝑟𝑟𝑟𝑟 and
𝐴𝐴𝑤𝑤 be denoted by 𝛹𝛹𝑟𝑟𝑟𝑟𝑟𝑟 (𝑠𝑠) and 𝛹𝛹𝑤𝑤 (𝑠𝑠), and the
least common multiple of 𝛹𝛹𝑟𝑟𝑟𝑟𝑟𝑟 (𝑠𝑠) and 𝛹𝛹𝑤𝑤 (𝑠𝑠) is
expressed as:

Assumption II: No root of 𝛹𝛹(𝑠𝑠) coincides with any


transmission zeros of the plant.
8

Servo-compensator Structure
The state space model of the servo-compensator
is given by:

Where 𝑥𝑥𝑐𝑐 ∈ 𝑅𝑅 𝑟𝑟ℎ and 𝐴𝐴𝑐𝑐 = 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏[𝐴𝐴𝑐𝑐1 , 𝐴𝐴𝑐𝑐2 ,…, 𝐴𝐴𝑐𝑐𝑟𝑟 ],


𝐵𝐵𝑐𝑐 = 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏[𝐵𝐵𝑐𝑐1 , 𝐵𝐵𝑐𝑐2 ,…, 𝐵𝐵𝑐𝑐𝑟𝑟 ], where
9

Servo-compensator Structure
Remark 1: Since (𝐴𝐴𝑐𝑐𝑖𝑖 , 𝐵𝐵𝑐𝑐𝑖𝑖 ) is controllable, (𝐴𝐴𝑐𝑐 , 𝐵𝐵𝑐𝑐 ) is
also controllable (why?).

Remark 2: This selection of servo-compensator


guarantees zero steady-state error for all
references and disturbances generated by:

Moreover, the proposed servo-compensator ensures


zero interactions between the control channels in
steady-state.
10

Servo-compensator: Example
Design a servo-compensator for the following
SISO system. w= 𝑎𝑎 sin(𝜔𝜔𝜔𝜔)
Stabilizing Compensator

yref =k.u-1(t) e Servo- xc


Kc
- u SISO y
compensator Plant
+
- -
xp
Kp

Solution:
0 𝜔𝜔
𝐴𝐴𝑟𝑟𝑟𝑟𝑟𝑟 = 0 𝐴𝐴𝑤𝑤 =
−𝜔𝜔 0
𝐶𝐶𝑟𝑟𝑟𝑟𝑟𝑟 = 𝑘𝑘 𝐶𝐶𝑤𝑤 = 1 1
11

Servo-compensator: Example

𝛹𝛹 𝑠𝑠 = 𝑠𝑠(𝑠𝑠 2 + 𝜔𝜔2 )

0 1 0 0
𝐴𝐴𝑐𝑐 = 0 0 1 , 𝑎𝑎𝑎𝑎𝑎𝑎 𝐵𝐵𝑐𝑐 = 0
0 −𝜔𝜔2 0 1
12

Control Problem Formulation


Define 𝑥𝑥 as: . Then, the model of
augmented closed-loop system is:

where
13

Control Objectives
The following objectives must be achieved:
• Closed-loop system should be AS.
• Coupling between control channels in response to
commands and disturbances should be minimized.
• Tracking error associated with the transient
response in all channels should be minimized.
To achieve the above objectives, the following cost
function is proposed.
14

Linear Quadratic Output Tracking


Define the linear transformation 𝛹𝛹 𝐷𝐷 where 𝐷𝐷
is the derivative operator. Apply 𝛹𝛹 𝐷𝐷 to state
space equations of the closed-loop system to
obtain:

where

Where 𝑧𝑧 = 𝛹𝛹 D 𝑥𝑥, υ = 𝛹𝛹 D u. Note that (why?):


15

Linear Quadratic Output Tracking


We can show that:

where:
16

Design of Stabilizing Compensator


To obtain 𝐾𝐾 = 𝐾𝐾𝑐𝑐 , 𝐾𝐾𝑝𝑝 , the following general LQR
cost function is defined and minimized.

where

Note that 𝑄𝑄(0) 𝑎𝑎𝑎𝑎𝑎𝑎 𝑅𝑅 should be positive-definite.


17

Design of Stabilizing Compensator


Note that the defined cost function is equal to:

where
18

Design Algorithm
19

Design Algorithm

Remark: Since the plant is controllable, a solution


to place the closed-loop poles in any desired
region always exists.
Examples and reference: M. Karimi-Ghartemani, S. A.
Khajehoddin, P. Jain, and A. Bakhshai, “Linear quadratic
output tracking and disturbance rejection,” International
Journal of control, vol. 84, no. 8, pp. 1442–1449, 2011.

You might also like