FUZZY SYSTEM MODELING BY FUZZY PARTITION AND GA HYBRID SCHEMES

This paper presents an approach to building multi-input and single-out put fuzzy models. Such a model is composed of fuzzy implications, and its output is inferred by simplified reasoning. The implications are a utomatically generated by the structure and parameter identification. In structure identification, the optimal or near optimal number of fuz zy implications is determined in view of valid partition of data set. The parameters defining the fuzzy implications are identified by a GA (Genetic Algorithm) hybrid scheme to minimize mean square errors globa lly. Numerical examples are provided to evaluate the feasibility of th e proposed approach. Comparison shows that the suggested approach can produce a fuzzy model with higher accuracy and a smaller number of fuz zy implications than the ones achieved previously in other methods. Th e proposed approach has also been applied to construct a fuzzy model f or the navigation control of a mobile robot. The validity of the resul tant model is demonstrated by experimentation. (C) 1997 Elsevier Scien ce B.V.