Abstract: Fuzzy systems have been successfully applied to the design of knowledge based controllers, yielding very good performance in many cases. However, fuzzy control still lacks general formal analysis and design techniques that allow the designer ensure a priori certain features of the closed-loop system, particularly stability. This article presents a simple systematic design procedure, based on approximate linearization, that guarantees closed-loop asymptotic stability. The method is applicable when a (nonfuzzy) plant model is available and a Takagi–Sugeno fuzzy controller is to be designed. The proposed design method is illustrated with an example and simulation results.
Abstract: This paper presents a new identification method for fuzzy models used in nonlinear prediction. The structure and parameters of the fuzzy model are obtained, using input-output data, by minimization of the prediction error. The predictive capacity of the fuzzy model is compared with other linear and non-linear models analyzing an illustrative example. The results show that the new method presents a better behavior.