Dynamic Modeling and Neural Network-Based Intelligent Control of Flexible Systems
Description
Comprehensive treatment of several representative flexible systems, ranging from dynamic modeling and intelligent control design through to stability analysis
Fully illustrated throughout, Dynamic Modeling and Neural Network-Based Intelligent Control of Flexible Systems proposes high-efficiency modeling methods and novel intelligent control strategies for several representative flexible systems developed by means of neural networks. It discusses tracking control of multi-link flexible manipulators, vibration control of flexible buildings under natural disasters, and fault-tolerant control of bionic flexible flapping-wing aircraft and addresses common challenges like external disturbances, dynamic uncertainties, output constraints, and actuator faults.
Expanding on its theoretical deliberations, the book includes many case studies demonstrating how the proposed approaches work in practice. Experimental investigations are carried out on Quanser Rotary Flexible Link, Quanser 2 DOF Serial Flexible Link, Quanser Active Mass Damper, and Quanser Smart Structure platforms.
The book starts by providing an overview of dynamic modeling and intelligent control of flexible systems, introducing several important issues, along with modeling and control methods of three typical flexible systems. Other topics include:
- Foundational mathematical preliminaries including the Hamilton principle, model discretization methods, Lagrange's equation method, and Lyapunov's stability theorem
- Dynamic modeling of a single-link flexible robotic manipulator and vibration control design for a string with the boundary time-varying output constraint
- Unknown time-varying disturbances, such as earthquakes and strong winds, and how to suppress them and use MATLAB and Quanser to verify effectiveness of a proposed control
- Adaptive vibration control methods for a single-floor building-like structure equipped with an active mass damper (AMD)
Dynamic Modeling and Neural Network-Based Intelligent Control of Flexible Systems is an invaluable resource for researchers and engineers seeking high-efficiency modeling methods and neural-network-based control solutions for flexible systems, along with industry engineers and researchers who are interested in control theory and applications and students in related programs of study.
More details
Other editions
Additional editions
Persons
Hejia Gao, PhD, is an Associate Professor at the School of Artificial Intelligence, Anhui University, Hefei, China. Previously, she was a Visiting Researcher at the Department of Mechanical, Industrial and Aerospace Engineering, Concordia University, Canada. She has published over 30 international journal and conference papers. Her research interests include neural networks, reinforcement learning, flexible systems, and vibration control.
Wei He, PhD, is a Full Professor at the School of Automation and Electrical Engineering, University of Science and Technology Beijing, China. He has co-authored three books and published over 100 international journal and conference papers. He was awarded a Newton Advanced Fellowship from the Royal Society, UK, in 2017. His research interests include adaptive control, vibration control, and bionic flapping wing aircraft.
Changyin Sun, PhD, is a Professor at the School of Automation, Southeast University, Nanjing, China. He has co-authored four books and published over 160 international journal papers. Prof. Sun is a Chinese Association of Automation Fellow. His research interests include intelligent control, flight control, pattern recognition, and optimal theory.
Content
About the Authors xi
Preface xiii
Acknowledgments xvii
Acronyms xix
1 Introduction 1
1.1 Background and Motivation 1
1.2 Modeling and Control Strategies of Flexible Robotic Manipulators 5
1.3 Vibration Control Technologies of Flexible Building-like Structures 7
1.4 Modeling and Control Approaches of Bionic Flexible Flapping-wing Aircraft 8
1.5 Outline of the Book 9
2 Mathematical Preliminaries 13
2.1 Mathematical Preliminaries 13
2.1.1 Hamilton Principle 13
2.1.2 Model Discretization 14
2.1.2.1 Assumed Mode Method 14
2.1.2.2 Finite Rigid Element Method 14
2.1.3 Lagrange's Equation Method 15
2.1.4 Neural Networks 15
2.1.5 Lyapunov Stability Theorem 16
2.1.6 Summary 18
3 Fuzzy Neural Network Control of the Single-Link Flexible Robotic Manipulator 19
3.1 Introduction 19
3.2 Problem Formulation 21
3.2.1 Dynamic Modeling 21
3.2.2 Model Discretization 22
3.3 Fuzzy Neural Network Control 24
3.3.1 Control Design 24
3.3.2 Stability Analysis 27
3.4 Numerical Simulations 30
3.4.1 Without Control 31
3.4.2 PD Control 32
3.4.3 Full-State Feedback 34
3.4.4 Output Feedback 34
3.5 Experimental Investigation 34
3.5.1 Experimental Testbed 34
3.5.2 Experimental Results 38
3.6 Summary 41
4 High-Gain Observer-Based Neural Network Control of the Two-Link Flexible Robotic Manipulator 43
4.1 Introduction 43
4.2 Problem Formulation 44
4.2.1 Dynamic Modeling 44
4.2.2 Model Discretization 47
4.3 High-Gain Observer-Based Neural Network Control 47
4.3.1 Control Design 47
4.3.2 Stability Analysis 49
4.4 Numerical Simulations 53
4.4.1 Simulation Results for Open-Loop System 53
4.4.2 Simulation Results for PD Control 54
4.4.3 Simulation Results for Neural Network Control 54
4.4.4 Comparison Between PD and NN Simulation Results 55
4.5 Experimental Investigation 58
4.5.1 Introduction of the Experimental Testbed 58
4.5.2 Experimental Results 58
4.5.3 Comparison Between PD and NN Experiment Results 61
4.6 Summary 62
5 Robust Adaptive Vibration Control for a String with Time-Varying Output Constraint 65
5.1 Introduction 65
5.2 Problem Formulation 67
5.2.1 Dynamics of the String System 67
5.2.2 Preliminaries 69
5.3 Control Design 69
5.3.1 Exact Model-Based Boundary Control 69
5.3.2 Robust Adaptive Boundary Control for System Parametric Uncertainty 72
5.4 The Solvability of the Inequality Equations 76
5.5 Numerical Simulations 81
5.6 Summary 84
6 Neural Network Vibration Control of a Stand-Alone Tall Building-Like Structure with an Eccentric Load 85
6.1 Introduction 85
6.2 Dynamic Modeling 88
6.2.1 Dynamic Modeling 88
6.2.2 Model Discretization 89
6.3 Neural Network Vibration Control 92
6.3.1 Control Design 92
6.3.2 Stability Analysis 93
6.4 Numerical Simulations 96
6.4.1 Simulation Parameters 96
6.4.2 Simulation Results 96
6.5 Experimental Investigation 100
6.5.1 Introduction of the Experimental Testbed 100
6.5.2 Experimental Results 101
6.6 Summary 105
7 Adaptive Vibration Control of a Flexible Structure Based on Hybrid Learning Controlled Active Mass Damping 107
7.1 Introduction 107
7.2 Dynamic Modeling 109
7.3 Hybrid Learning Control 113
7.3.1 Disturbance Observer Design 113
7.3.2 Hybrid Learning Control Design 115
7.3.3 Full-order State Observer 118
7.4 Simulation Verification and Comparative Analysis 118
7.5 Experimental Investigation 120
7.5.1 Experimental Results of Passive Mode 122
7.5.2 Experimental Results of PV Position Controller 124
7.5.3 Experimental Results of HL Controller 125
7.5.4 Comparisons and Discussions 128
7.6 Summary 129
8 Reinforcement Learning Control of a Single-Floor Building-Like Structure with Active Mass Damper 131
8.1 Introduction 131
8.2 Problem Formulation 132
8.2.1 Dynamic Modeling 132
8.2.2 Model Discretization 134
8.3 Reinforcement Learning Control 134
8.3.1 Control Design 134
8.3.2 Stability Analysis 136
8.4 Experimental Investigation 137
8.5 Summary 141
9 Disturbance Observer-Based Neural Network Control of a Flexible Flapping-Wing System 143
9.1 Introduction 143
9.2 Problem Formulation 144
9.2.1 Dynamic Modeling 144
9.2.2 Model Discretization 146
9.3 Disturbance Observer-Based Neural Network Control 148
9.3.1 Control Design 148
9.3.2 Stability Analysis 152
9.3.3 Simulation Results Without Control 155
9.3.4 Simulation Results for PD Control 155
9.3.5 Simulation Results for Full-State Feedback 155
9.3.6 Simulation Results for Output Feedback 158
9.4 Summary 159
10 Adaptive Finite-Time Control of a Bionic Flexible Flapping-Wing Aircraft with Actuator Failures 161
10.1 Introduction 161
10.2 Problem Formulation 163
10.2.1 Dynamic Modeling 164
10.2.2 Model Discretization 165
10.3 Adaptive Finite-Time Control 167
10.3.1 Control Design 167
10.3.2 Stability Analysis 169
10.4 Numerical Simulations 172
10.5 Summary 181
11 Adaptive Vibration Control for Two-Stage Bionic Flapping Wings Based on Neural Network Algorithm 183
11.1 Introduction 183
11.2 Problem Formulation 184
11.2.1 Dynamic Modeling 184
11.2.2 Model Discretization 185
11.3 Adaptive Vibration Control 186
11.3.1 Control Design 186
11.3.2 Stability Analysis 188
11.4 Numerical Simulations 191
11.5 Summary 195
12 Boundary Vibration Control of a Floating Wind Turbine System with Mooring Lines 197
12.1 Introduction 197
12.2 System Modeling and Preliminaries 199
12.2.1 Dynamical Model of Floating Wind Turbine Vibrations 200
12.2.2 Preliminaries 201
12.3 Controller Design 202
12.4 Numerical Simulations 206
12.5 Summary 215
13 Conclusions 217
References 219
Index 243
Preface
The flexible system covers many different objects such as flexible robotic manipulators, bionic flexible flapping wing aircraft, and flexible buildings. With a large number of applications of flexible systems, its control theory and method issues have become a prospective high-tech research direction, which attracts concerns from both academic and industrial fields. At present, the control theory and method of flexible systems, such as the tracking and vibration control of multi-link flexible manipulators, the constraint control of flexible buildings under natural disasters, and the fault-tolerant control of bionic flexible flapping-wing robots, has developed into a common scientific problem, which is extremely challenging. In order to solve the technical problems of dynamic modeling and intelligent control of uncertain flexible systems with environmental adaptability, the book makes a systematic and detailed study on modeling mechanism and control strategy of several flexible systems.
Chapter 1 provides an overview of dynamic modeling and intelligent control of flexible systems, introducing several important issues in the study of flexible systems. The modeling and control methods of three typical flexible systems are discussed separately.
Chapter 2 provides the corresponding mathematical preliminaries of subsequent chapters, including the Hamilton principle, model discretization methods, Lagrange's equation method, neural networks, and Lyapunov stability theorem.
Chapter 3 develops the dynamic model of the single-link flexible robotic manipulator, which overcomes the challenge from the system dynamics being infinite dimensional. The fuzzy neural network control with uniform approximation performance is designed to solve the system uncertainties. Numerical simulations and extensive experiments have been investigated to verify the effectiveness of the proposed methods.
Chapter 4 establishes a finite-dimensional dynamic model of the two-link flexible robotic manipulator. A high-gain observer-based neural network control strategy is proposed to estimate the immeasurable states in practice. The semi-globally uniformly ultimate boundedness (SGUUB) of the closed-loop system is guaranteed via Lyapunov's stability theory. The simulation and experimental results demonstrate the effectiveness of the proposed control strategy.
Chapter 5 present the vibration control design for a string with the boundary time-varying output constraint. The dynamics of the string is a distributed parameter system described by a partial differential equation and two ordinary differential equations. A barrier Lyapunov function with a logarithmic function is adopted to prevent the time-varying constraint violations. Adaptive control is designed to handle the system parametric uncertainties. Stability analysis and the solvability of the inequality equations are provided. Numerical simulations are provided to illustrate the effectiveness of the proposed control design.
Chapter 6 focuses on a stand-alone tall building-like structure with an eccentric load. A neural network control approach is proposed to suppress vibrations caused by unknown time-varying disturbances (earthquake, strong wind, etc.). The output constraint on the angle of the eccentric load is also considered, and such angle can be ensured within the safety limit by incorporating a barrier Lyapunov function. Simulations and experiments based on MATLAB and Quanser are carried out to verify the feasibility and effectiveness of the proposed control.
Chapter 7 discusses an adaptive vibration control method for a single-floor building-like structure equipped with an active mass damper (AMD). The method uses a hybrid learning control strategy to suppress vibrations caused by unknown time-varying disturbances such as earthquakes or strong winds. The effectiveness of the proposed control approach is demonstrated through experimental investigation on a Quanser Active Mass Damper. The research results aim to bring new ideas and methods to the field of disaster reduction for engineering development.
Chapter 8 investigates a single-floor building-like structure equipped with an active mass damper (AMD). Optimal vibration control, while dealing with system uncertainties, is realized by the reinforcement learning technique. When the unexpected natural disasters occur, the proposed controller applying to the active mass damper can compensate the increase of the system vibration caused by external disturbances. The experimental results in the form of graphics and tables have shown the effectiveness of the proposed control algorithm.
Chapter 9 develops the visualization model of the rigid-flexible coupled bionic flapping wing by the advanced system-level modeling software MapleSim. A novel neural network controller based on disturbance observer technology is proposed to compensate the system uncertainties. The proposed method can successfully suppress the vibration of the flapping wing while accurately track the desired trajectory. Co-simulation results from MapleSim and Matlab/Simulink validate the effectiveness of the proposed method.
Chapter 10 focuses on the flexible wings of the aircraft, which has great advantages, such as being lightweight, having high flexibility, and offering low energy consumption. A novel adaptive finite-time controller based on the fuzzy neural network and nonsingular fast terminal slidingmode scheme are proposed for tracking control and vibration suppression of the flexible wings, while successfully addressing the issues of system uncertainties and actuator failures. Co-simulations through MapleSim and MATLAB/Simulink are carried out to verify the performance of the proposed controller.
Chapter 11 discusses the importance of vibration control for bionic flapping-wing robotic aircraft and autonomous ornithopter applications. A visualization model of the rigid-flexible coupled bionic flapping wing is established using MapleSim software. A novel adaptive vibration controller based on neural network (NN) algorithm is proposed to compensate for system uncertainties. The proposed method can successfully suppress the vibration of the flapping wing while accurately tracking the desired trajectory. The effectiveness of the proposed method is validated through co-simulation results from MapleSim and Matlab/Simulink.
Chapter 12 investigate dynamic modeling, active boundary control design, and stability analysis for a coupled floating wind turbine (FWT) system, which is connected with two flexible mooring lines. It is a coupled beam-strings structure, and we design two boundary controllers to restrain the vibrations of this flexible system caused by external disturbances based on the coupled partial differential equations and ordinary differential equations (PDEs-ODEs) model. Meanwhile, significant performance of designed boundary controllers and system's stability are theoretically analyzed, and a set of simulation results are provided to show efficacy of the proposed approach.
Chapter 13 summarizes the practical significance in the application of neural network-based intelligent control and proposes some future research directions in this field.
In summary, this book proposes high-efficiency modeling methods and novel intelligent control strategies for several representative flexible systems developed by means of neural networks. The book discusses the tracking control of multi-link flexible manipulators, the vibration control of flexible buildings under natural disasters, and the fault-tolerant control of bionic flexible flapping-wing aircraft. Expanding on its theoretical deliberations, the book includes many case studies demonstrating how the proposed approaches work in practice. The most important features of the book include:
- a comprehensive review of modeling and control theory for flexible systems;
- detailed presentation of the modeling methods and the neural network-based control strategies;
- successful addressing of external disturbances, dynamic uncertainties, output constrains, and actuator faults;
- abundance case studies illustrating the important steps in designing the neural network-based control; and
- performance analysis of the described control approaches by a large number of figures and tables.
This book can be regarded as an authoritative reference for the field (studies) of dynamics and control of flexible systems. Interested readers will gain a systematic understanding of the flexible systems as well as the technical details involved. The material presented in this book will be useful for researchers and engineers who wish to avoid excessive time in searching high-efficiency modeling methods and neural-network-based control solutions for flexible systems. It is written for industry engineers and researchers who are interested in control theory and the applications. This book is also good for postgraduate students engaged in self-study of adaptive control for the flexible systems.
Hejia GaoAnhui University, China
Wei He
University of Science and Technology Beijing, China
Changyin Sun
Southeast University, China...
