Adaptive Scheme Based On Fuzzy Systems And Its Improving

Identification and control of nonlinear systems is a difficult but important and interrelated problem. Various nonlinear model structures such as polynomials, splines, wavelets, neural networks and fuzzy systems have been used to represent the nonlinear mechanisms in the nonlinear systems in the literature. Fuzzy system models, compared with other schemes, have the advantages of incorporating human know- ledge, being easy-to-understand and easy-to-implement, and fast convergence due to the convenience in parameter initialization.This dissertation focuses on identification and control of the nonlinear systems by fuzzy systems, and presents the analyses and improving on some of the involved adaptive methods. The simulation results show the validity of these methods.The fuzzy system is composed of the product inference engine, fuzzifier and defuz- zifier. This paper mostly researches on the adaptive(self-tuning) methods which are included in the identification of nonlinear systems and control of some nonlinear systems. The detailed is as follows:Firstly, we improve on the fuzzy implication, which derives from the fuzzy rule base, by introducing the compensatory algorithm. This reform will give the better results on approximation accuracy for the nonlinear systems.Secondly, we present a new adaptive fuzzy tracking system based on errors compensation, the method can availably reduce the tracking errors. With the same configuration of fuzzy system as referred above, we construct the fuzzy system with Gauss fuzzifier, and this new fuzzy system has the obvious strongpoints, it not only can overcome the input noises, but also has the robustness for the time-delay system.Thirdly, on the question of input variable membership division about fuzzy system, we introduce an idea of variable universe, which makes the control results of fuzzy systems improved.Fourthly, combining input-output linearization with adaptive fuzzy control, we present an improving scheme about adaptive fuzzy control method based on the input-output linearization.Finally, we discuss qualitatively the structural parameters of the adaptive fuzzy control systems, and present some methods which make the control effect better.To every improved scheme as referred above, we all present the simulation results, which show the validity of those methods. With the further development of the theory on fuzzy systems, we believe that the applied fields of fuzzy systems will expand continuously and the applied level will improve increasingly.