Synergy of switched-fuzzy and fuzzy-neural nonlinear systems enhances complexity and potential

Abstract

In this paper we present concepts for synergy of switched fuzzy and fuzzy-neural systems. First, an algorithm/procedure for neural network identification of switched fuzzy models, out of input-output data pairs is given. In order to use the existing stability and stabilization results in the field of switched fuzzy systems to the identified switched fuzzyneural models, an extension of the switched fuzzy model with levels of structure is presented. The proposed concepts are used for identification of discrete switched fuzzy models. To confirm the proposed algorithm/procedure, and the new extended model, a fuzzy-neural identification of the discrete switched fuzzy model for the nonholonomic WMR vehicle is presented. Based on the identified discrete switched fuzzy model for the WMR vehicle, design of discrete switched fuzzy controller is made. The simulation results show the effectiveness of the proposed concepts.