Fuzzy control of nonlinear systems by learning method and frequency response
Carlos Alberto Murari Pinheiro, Fernando A. C. Gomide · 1999
The purpose of this paper is to introduce a methodology to design nonlinear control systems using fuzzy controllers, a learning method, and frequency response information. It is well known that in nonlinear control systems, the dynamic response depends of the input intensity. The methodology proposed provides fuzzy controllers that produce system responses consistent with design requirements and less sensitive to input intensity. The work introduces the use of frequency response information together with a learning approach to design a class of fuzzy controllers with the proper membership functions and rules. The method exploits solid control engineering practice and notions to provide efficient solutions and conceptual links between fuzzy and classical control theory.