Adaptive neuro-fuzzy identification method of hammerstein model

Abstract

In this paper, adaptive neuro-fuzzy identification is investigated for the Hammerstein model, which consists of the cascade structure of a static nonlinearity followed by a linear dynamic part. Utilizing the approximation ability of neuro-fuzzy for the nonlinear static function, there is no need for prior knowledge and restriction on static nonlinear function. Furthermore, an adaptive algorithm designed by Lyapunov stability theory is proposed to obtain the neuro-fuzzy Hammerstein model. Example is used to illustrate the performance and applicability of the proposed neuro-fuzzy Hammerstein model.