Stability analysis of fuzzy-model-based control systems. Linear-matrix-inequality approach. (Q625124)

This specialized volume addresses an important and timely topic of a thorough stability analysis of fuzzy models. It falls under a rubric of a systematic development of control strategies based upon fuzzy rule-based models. From a historical perspective, fuzzy control (such as the one encountered in fuzzy controllers) was predominantly based upon domain knowledge of experts: neither a model nor a performance index were provided and used in the design of control policies in an explicit manner. In contrast, in further investigations, fuzzy rule-based models were involved along with some associated performance index capturing the quality of control itself. In the book, rule-based models expressed in the form of ``\(N\)'' rules ``if \(f_1({\mathbf x}(t))\) is \(A_{i1}\) and \(\dots\) and \(f_n({\mathbf x}(t))\) is \(A_{in}\) then \(d{\mathbf x}(t)/dt= A_i{\mathbf x}(t)+ B_i{\mathbf u}(t)\)'', \(i= 1,2,\dots, N\) with \(A_{i1},\dots, A_{in}\) being fuzzy sets associated with the corresponding functions of the state variables are studied. Likewise a control strategy is captured through a series of control rules containing linear state feedback sub-controllers. The stability analysis is inherently based on the Linear Matrix Inequality (LMI) approach. The material of the book is arranged in 10 chapters. Following a brief review of the background material (Chapter 1), stability and performance conditions for state feedback control systems are discussed in Chapter 2. The consecutive chapters (Chapter 3--6) focus on more specialized scenarios of stability analysis including membership function shape dependent stability, time varying feedback (invoking bilinear matrix inequalities), parameter-dependent Lyapunov functions, and regional switching fuzzy controllers. Chapter 7 tackles a problem of combining the state feedback fuzzy controller and switching controller and investigates its stability. The discussion covered in Chapter 8 brings an issue of time delay present in system states along with the ensuing stability analysis. Model reference tracking control is covered in Chapter 9. Further extensions to fuzzy models with type-2 fuzzy sets (membership functions) are presented in Chapter 10. An important problem, which arises from a practical perspective, is the linkage between the fuzzy model (and the membership functions existing in the rules) and the quality of control and its properties (such as stability) versus the control of real-world system being modeled through rules. Obviously, the same applies to control problems based on ``classic'' numerical models, however in fuzzy models and fuzzy control the influence of the linguistic description of a system under control, the impact of semantics (membership functions). and the level of the granularity of the model on the stability of control calls for more investigations and further development of practical guidelines. The book tackles some aspects of such fundamental design issues and offers interesting results. All in all, the volume is an indispensable source of knowledge for those interested in fuzzy control and control engineering, in general.