ACTIVE DISTURBANCE

REJECTION CONTROL
FOR NONLINEAR
SYSTEMS
ACTIVE DISTURBANCE
REJECTION CONTROL
FOR NONLINEAR
SYSTEMS
AN INTRODUCTION

Bao-Zhu Guo
Academy of Mathematics and Systems Science, Academia Sinica, People’s Republic
of China and University of the Witwatersrand, South Africa

Zhi-Liang Zhao
Shaanxi Normal University, Xi’an, People’s Republic of China
This edition first published 2016 © 2016 John Wiley & Sons (Asia) Pte Ltd.
Registered office
John Wiley & Sons Singapore Pte. Ltd., 1 Fusionopolis Walk, #07-01 Solaris South Tower, Singapore
138628.
For details of our global editorial offices, for customer services and for information about how to apply
for permission to reuse the copyright material in this book please see our website at www.wiley.com.
All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, scanning,
or otherwise, except as expressly permitted by law, without either the prior written permission of the
Publisher, or authorization through payment of the appropriate photocopy fee to the Copyright
Clearance Center. Requests for permission should be addressed to the Publisher, John Wiley & Sons
Singapore Pte. Ltd., 1 Fusionopolis Walk, #07-01 Solaris South Tower, Singapore 138628, tel:
65-66438000, fax: 65-66438008, email: [email protected].
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print
may not be available in electronic books.
Designations used by companies to distinguish their products are often claimed as trademarks. All
brand names and product names used in this book are trade names, service marks, trademarks or
registered trademarks of their respective owners. The Publisher is not associated with any product or
vendor mentioned in this book. This publication is designed to provide accurate and authoritative
information in regard to the subject matter covered. It is sold on the understanding that the Publisher is
not engaged in rendering professional services. If professional advice or other expert assistance is
required, the services of a competent professional should be sought.
Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts
in preparing this book, they make no representations or warranties with respect to the accuracy or
completeness of the contents of this book and specifically disclaim any implied warranties of
merchantability or fitness for a particular purpose. It is sold on the understanding that the publisher is
not engaged in rendering professional services and neither the publisher nor the author shall be liable
for damages arising here from. If professional advice or other expert assistance is required, the services
of a competent professional should be sought.
Library of Congress Cataloging-in-Publication Data
Names: Guo, Bao-Zhu, 1962- author. | Zhao, Zhi-Liang, author.
Title: Active disturbance rejection control for nonlinear systems : an
introduction / Bao-Zhu Guo and Zhi-Liang Zhao.
Description: Singapore ; Hoboken, NJ : John Wiley & Sons, 2016. | Includes
bibliographical references and index.
Identifiers: LCCN 2016018239 (print) | LCCN 2016025091 (ebook) | ISBN
9781119239925 (cloth) | ISBN 9781119239956 (pdf) | ISBN 9781119239949
(epub)
Subjects: LCSH: Damping (Mechanics) | Automatic control. | Nonlinear systems.
Classification: LCC TA355 .G79 2016 (print) | LCC TA355 (ebook) | DDC
620.3/7–dc23
LC record available at https://lccn.loc.gov/2016018239
A catalogue record for this book is available from the British Library.
Set in 10/12pt, TimesLTStd by SPi Global, Chennai, India.

1 2016
Contents

Preface ix

1 Introduction 1
1.1 Problem Statement 1
1.2 Overview of Engineering Applications 6
1.3 Preliminaries 9
1.3.1 Canonical Form of ADRC 9
1.3.2 Stability for Nonlinear Systems 18
1.3.3 Stability of Linear Systems 24
1.3.4 Finite-Time Stability of Continuous System 27
1.3.5 Stability of Discontinuous Systems 32
1.3.6 Proof of Theorem 1.3.11 34
1.4 Remarks and Bibliographical Notes 50

2 The Tracking Differentiator (TD) 53

2.1 Linear Tracking Differentiator 55
2.2 Nonlinear Tracking Differentiator 59
2.2.1 Second-Order Nonlinear Tracking Differentiator 60
2.2.2 High-Order Nonlinear Tracking Differentiator 64
2.3 Finite-Time Stable System-Based Tracking Differentiator 69
2.3.1 Convergence of Finite-Time Stable System-Based TD 70
2.3.2 A Second-Order Finite-Time Stable Tracking Differentiator 75
2.4 Illustrative Examples and Applications 77
2.4.1 Comparison of Three Differentiators 77
2.4.2 Applications to Frequency Online Estimation 81
2.4.3 Application to the Boundary Stabilization of Wave Equation 85
2.5 Summary and Open Problems 88
2.6 Remarks and Bibliographical Notes 89
vi Contents

3 Extended State Observer 93

3.1 Linear Extended State Observer for SISO Systems 94
3.2 Nonlinear Extended State Observer for SISO Systems 100
3.2.1 Nonlinear ESO for SISO Systems 101
3.2.2 Some Special ESO 109
3.3 The ESO for SISO Systems with Time-Varying Gain 119
3.4 The ESO for MIMO Systems with Uncertainty 133
3.4.1 ESO for Systems with Total Disturbance 133
3.4.2 ESO for Systems with External Disturbance Only 141
3.4.3 Examples and Numerical Simulations 144
3.5 Summary and Open Problems 150
3.6 Remarks and Bibliographical Notes 153

4 The Active Disturbance Rejection Control 155

4.1 Linear ADRC for SISO Systems 157
4.1.1 Global Convergence of Linear ADRC for SISO Systems 157
4.1.2 Global Convergence for Systems with External Disturbance Only 167
4.1.3 Semi-Global Convergence of LADRC 176
4.1.4 Numerical Simulations 183
4.2 Nonlinear ADRC for SISO Systems 187
4.2.1 Global ADRC for SISO Systems with Total Disturbance 187
4.2.2 Global ADRC for SISO System with External Disturbance Only 195
4.2.3 Semi-global ADRC for SISO System with Vast Uncertainty 203
4.2.4 Examples and Numerical Simulations 204
4.3 ADRC with Time-Varying Tuning Parameter 206
4.4 Nonlinear ADRC for MIMO Systems with Vast Uncertainty 228
4.4.1 Semi-Global ADRC for MIMO Systems with Uncertainty 231
4.4.2 Global ADRC for MIMO Systems with Uncertainty 238
4.4.3 Global ADRC for MIMO Systems with External Disturbance Only 245
4.4.4 Numerical Simulations 250
4.5 IMP Versus ADRC 252
4.6 HGC and SMC Versus ADRC 263
4.7 Applications to PMSMs 267
4.8 Application to Wave Equation with Uncertainty 270
4.8.1 Control Design 271
4.8.2 Proof of Theorem 4.8.1 279
4.9 Summary and Open Problems 287
4.10 Remarks and Bibliographical Notes 289

5 ADRC for Lower Triangular Nonlinear Systems 291

5.1 ESO for Lower Triangular Systems 291
5.1.1 Constant High-Gain ESO 292
5.1.2 Time-Varying Gain ESO 301
5.1.3 Numerical Simulation 306
Contents vii

5.2 Stabilization of Lower Triangular Systems by ADRC 312

5.2.1 ADRC with Constant Gain ESO 313
5.2.2 ADRC with Time-Varying Gain ESO 327
5.3 Numerical Simulations 331
5.4 Summary and Open Problems 336
5.5 Remarks and Bibliographical Notes 338

References 341

Index 349
Preface

The modern control theory came to mathematics via N. Wiener’s book Cybernetics or the
Science of Control and Communication in the Animal and the Machine, published in 1948.
In 1954, H.S. Tsien published the book Engineering Cybernetics, which brought the control
theory to engineering.
Roughly speaking, the modern control theory has been developed through three stages. The
first stage is the classical control or automatic principle of compensation developed from the
1940s to the 1960s. During this period the single-input and single-output linear time-invariant
systems were studied by the frequency domain approach. The second stage was from the 1960s
to the 1980s, during which time the multi-input and multi-output systems were studied by the
state space or time domain approach. The state-space approach relies heavily upon the math-
ematical models of the systems. After the 1980s, many control theories were developed to
cope with uncertainties in systems. Several powerful methods were developed, including the
internal model principle for output regulation (which started in the 1970s), as well as adaptive
control, roust control, high-gain feedback control, and sliding mode control (which started
even earlier). In particular, the robust control theory was well-established by both the fre-
quency domain approach and the time domain approach. A common feature for these methods
was the worst case scenario regarding the disturbance. A different way of dealing with uncer-
tainty may be found in adaptive control where the unknown parameters are estimated under
the “exciting persistent condition” and in output regulation where a special class of external
disturbance is estimated through the observer and internal model and is compensated for in
the feedback-loop.
During the late 1980s and 1990s, Jingqing Han of the Chinese Academy of Sciences pro-
posed a powerful unconventional control approach to deal with vast uncertainty in nonlinear
systems. This new control technology was later called the active disturbance rejection con-
trol (ADRC). The uncertainties dealt with by the ADRC can be very complicated. They could
include the coupling of the external disturbances, the system unmodeled dynamics, the zero
dynamics with unknown model, and the superadded unknown part of control input. The key
idea of the ADRC considers the “total disturbance” as a signal of time, which can be estimated
by the output of the system. Basically, the ADRC consists of three main parts. The first part is
the tracking differentiator (TD) that is relatively independent and is actually thoroughly dis-
cussed in the control theory. The aim of the TD is to extract the derivatives of the reference
signal and is also considered as transient profile for output tracking. The second part of the
ADRC is the extended state observer (ESO) which is a crucial part of the ADRC. In ESO, both
the state and the “total disturbance” are estimated by the output of the system. This remarkable
x Preface

feature makes the ADRC a very different way of dealing with uncertainty. The ESO is the
generalization of the traditional state observer where only the state of the system is estimated.
The final part of the ADRC is the extended state observer-based feedback control. Since the
uncertainty is estimated in the ESO and is compensated for in the feedback loop, the barriers
between the time invariant and time varying, linear and nonlinear have been broken down by
considering the time-varying part and the nonlinear part as uncertainty. At the same time, the
control energy is significantly reduced. More importantly, in this way, the closed-loop systems
look like linear time-invariant systems, for which a reliable result can be applied.
In the past two decades, the ADRC has been successfully applied to many engineering
control problems such as hysteresis compensation, high pointing accuracy and rotation speed,
noncircular machining, fault diagnosis, high-performance motion control, chemical processes,
vibrational MEMS gyroscopes, tension and velocity regulations in Web processing lines and
DC–DC power converters by many researchers in different contexts. In all applications of
process control and motion control, compared with the huge amount of literature on control
theory in dealing with the uncertainty such as system unmodeled dynamics, external distur-
bance rejection, and unknown parameters, the ADRC has shown its remarkable PID nature of
an almost independent mathematical model, no matter the high accuracy control of micrometre
grade or the integrated control on a very large scale.
On the other hand, although many successful engineering applications have been developed,
the theoretical research on these applications lags behind. This book serves as an introduction
to the ADRC from a theoretical perspective in a self-contained way. In Chapter 1, some basic
background is introduced on nonlinear uncertain systems that can be dealt with by the ADRC.
Chapter 2 presents convergence of the different types of tracking differentiators proposed by
Han in his original papers. Chapter 3 is devoted to convergence of the extended state observer
for various nonlinear systems. Chapters 2 and 3 can be considered as independent sections of
the book. Chapter 4 looks at convergence of the closed-loop based on the TD and ESO. This
can be considered as a separation principle of the ADRC for uncertain nonlinear systems.
The numerical simulations are presented from here to where to illustrate the applicability of
the ADRC. Finally, in Chapter 5, the ESO and stabilization for lower triangular systems are
discussed.
Most of the material in this book is from the authors’ published papers on this topic.
However, the idea for the book comes from ideas published in Han’s many original numerical
experiments, engineering applications that appeared publicly in Chinese, and the numerous
works done by the ADRC group, in particular the group led by Dr. Zhiqiang Gao at Cleveland
State University in the United States.
The authors are deeply indebted to those who helped with the works presented in this book.
These include Zhiqiang Gao, Yi Huang, and Wenchao Xue. This book is dedicated to Bao-Zhu
Guo’s memory of Professor Jingqing Han who passed away in 2008.
This book is partially supported by the National Natural Science Foundation of China
(No. 61403242) and the Nature Science Research Program of Shaanxi Province-Key Program
(No. 2016JZ023).

Bao-Zhu Guo and Zhi-Liang Zhao

October, 2015