Neuro-Fuzzy Control of Industrial Systems With Actuator Nonlinearities

by Lewis, Frank L.; Campos, Javier; Selmic, Rastko

• ISBN13: 9780898715057
• ISBN10: 0898715059
• Format: Hardcover
• Copyright: 2001-12-01
• Publisher: Society for Industrial & Applied
• More Book Details
• Purchase Benefits

Summary

Neural networks and fuzzy systems are model free control design approaches that represent an advantage over classical control when dealing with complicated nonlinear actuator dynamics. Neuro-Fuzzy Control of Industrial Systems with Actuator Nonlinearities brings neural networks and fuzzy logic together with dynamical control systems. Each chapter presents powerful control approaches for the design of intelligent controllers to compensate for actuator nonlinearities such as time delay, friction, deadzone, and backlash that can be found in all industrial motion systems, plus a thorough development, rigorous stability proofs, and simulation examples for each design. In the final chapter, the authors develop a framework to implement intelligent control schemes on actual systems. Rigorous stability proofs are further verified by computer simulations, and appendices contain the computer code needed to build intelligent controllers for real-time applications.

Table of Contents

Preface	xiii
Background on Neural Networks and Fuzzy Logic Systems	1	(12)
Neural networks	1	(7)
Two-layer neural network topology and tuning	1	(3)
Single-layer neural networks	4	(2)
Universal approximation property of neural networks	6	(2)
Fuzzy logic systems	8	(5)
Membership functions and defuzzification	9	(1)
Universal approximation property and tuning of fuzzy logic networks	10	(3)
Background on Dynamical Systems and Industrial Actuators	13	(16)
Dynamical systems	13	(8)
Continuous and discrete dynamical systems	13	(2)
Mathematics and system properties	15	(3)
Lyapunov stability analysis	18	(3)
Industrial actuators	21	(1)
Friction	22	(2)
Deadzone	24	(2)
Backlash	26	(3)
Neurocontrol of Systems with Friction	29	(24)
Neural network approximation of functions with discontinuities	29	(10)
Neural network approximation of continuous functions	30	(1)
Neural network approximation of functions with jumps	30	(5)
Augmented multilayer neural networks for jump function approximation	35	(1)
Simulation of the augmented multilayer neural network	36	(3)
Neurocontroller for friction compensation	39	(11)
Robot arm dynamics	40	(2)
Neural network controller	42	(6)
Simulation of a neural network controller	48	(2)
Conclusions	50	(3)
Neural and Fuzzy Control of Systems with Deadzones	53	(38)
Introduction to deadzone control	53	(1)
Position tracking controller with neural network deadzone compensation	54	(16)
Neural network deadzone precompensator	54	(6)
Robot arm dynamics	60	(1)
Design of tracking controller with neural network deadzone compensation	61	(6)
Simulation of neural network deadzone compensator for robotic systems	67	(3)
Fuzzy logic discrete-time deadzone precompensation	70	(13)
Fuzzy logic discrete-time deadzone precompensator	73	(4)
Discrete-time fuzzy logic controller with deadzone compensation	77	(4)
Simulation results	81	(2)
Appendix: Proof of Theorem 4.3.1	83	(8)
Neural Control of Systems with Backlash	91	(34)
Introduction to backlash control	91	(1)
Continuous-time neural network backlash compensation	92	(13)
Backlash inverse	93	(1)
Dynamics and control of nonlinear motion systems	94	(2)
Neural network backlash compensation using dynamic inversion	96	(7)
Simulation of neural network backlash compensator	103	(2)
Discrete-time neural network backlash compensation	105	(10)
Discrete-time backlash nonlinearity and backlash inverse	107	(1)
Dynamics and control of a nonlinear discrete-time system	108	(2)
Discrete-time neural network backlash compensation using dynamic inversion	110	(3)
Simulation results	113	(2)
Appendix: Proof of Theorem 5.3.2	115	(10)
Fuzzy Logic Control of Vehicle Active Suspension	125	(14)
Introduction	125	(2)
System model and dynamics	127	(3)
Backstepping-based fuzzy logic controller	130	(6)
Outer loop design for ideal force input	130	(3)
Inner backstepping loop design	133	(3)
Simulation results	136	(3)
Neurocontrol Using the Adaptive Critic Architecture	139	(12)
Introduction	139	(2)
Dynamics of a nonlinear system	141	(1)
Adaptive critic feedback controller	142	(3)
Adaptive critic neurocontrol architecture	142	(3)
Simulation results	145	(2)
Conclusions	147	(2)
Appendix: Proof of Theorem 7.3.1	149	(2)
Neurocontrol of Telerobotic Systems with Time Delays	151	(20)
Introduction	151	(2)
Background and problem statement	153	(2)
Identification of time-delay-free systems using recurrent neural networks	155	(2)
Recurrent neural network predictive control strategy for compensating time delays	157	(3)
Review of the Smith predictor	158	(1)
Control structure	159	(1)
Recurrent neural network predictive control and stability analysis	160	(3)
Simulation example	163	(3)
Conclusion	166	(1)
Appendix: Proof of Theorem 8.3.1	167	(1)
Appendix: Proof of Theorem 8.5.1	168	(3)
Implementation of Neural Network Control Systems	171	(16)
PC-PC real-time digital control system	171	(7)
Hardware description	172	(2)
Software description	174	(2)
State-variable data management	176	(2)
ATB1000 U.S. Army tank gun barrel testbed	178	(1)
Derivation of neural net control system for flexible systems	179	(4)
Flexible-link robot dynamics	180	(1)
Singular perturbation approach	181	(1)
Neural network control algorithm	182	(1)
Implementation of neural network controller on ATB1000 testbed	183	(3)
Proportional-plus-derivative control	185	(1)
Proportional-plus-derivative control plus neural network	185	(1)
Conclusions	186	(1)
Appendix A C Code for Neural Network Friction Controller	187	(14)
Appendix B C Code for Continuous-Time Neural Network Deadzone Controller	201	(18)
Appendix C C Code for Discrete-Time Neural Network Backlash Controller	219	(10)
Appendix D Versatile Real-Time Executive Code for Implementation of Neural Network Backstepping Controller on ATB1000 Tank Gun Barrel	229	(4)
References	233	(8)
Index	241