The book is devoted to intelligent control systems characterized by adaptation, learning, autonomy, and complexity. The authors approaches to intelligent control are based on adaptive control theory and concepts of artificial intelligence. The historical background of these originally independent scientific branches is presented in the introduction. For dealing with complexity temporal and spatial multimodality is used while uncertainty is handled with stochastic adaptive control.
The second part of the book is devoted to deterministic systems. Adaptive control of nonlinear systems is presented for continuous-time and discrete-time systems including the well-known feedback linearization. Dynamic structure networks allow stable adaptive control, too. The corresponding control law and error dynamics are presented as well as a stability analysis. The composite adaptive control of continuous-time systems uses neural networks where the control law is supplemented by the adaptive law. Finally, the functional adaptive control of discrete-time systems is considered using neural networks again. Further, an extension to adaptive sliding mode control is made.
The third part of the book deals with stochastic systems. The fundamental principles and two problem classes are defined: known state-space models with inaccessible states and statespace or input-output models, respectively, with unknown parameters. Then, dual adaptive control of nonlinear systems is introduced. The multimodal approaches are proposed for systems with unknown parameters. The multiple model dual adaptive control is treated in detail for jump nonlinear systems and spatial multimodal systems.
The authors present a complete survey on intelligent control. All the methods are well explained and many simulation examples show the performance of the resulting control. A comprehensive list of references is also included. The book is recommended to all control engineers and system dynamicists who are dealing with complex systems. The presented methods offer remarkable improvements for nonlinear systems with unknown or time-varying parameters.
Stuttgart W. Schiehlen