Approximation Error Bounds of Minimum Inference Fuzzy System as Function Approximator

摘要

In this paper, based on a minimum inference fuzzy system defuzzified by the center of average defuzzifier and linear expression of the fuzzy system, problems of approximation error bounds are discussed by linear interpolation theory. Approximation error bounds of minimum inference fuzzy system of MISO with local properties are given. Error remainder terms and auxiliary functions have been used to the study of the approximation error bounds. New insight on interpolation capabilities of the fuzzy system is obtained by this method. Numeric example is given to illustrate effectiveness of the approximation error bounds.