A novel identification method for Takagi–Sugeno fuzzy model

Abstract

Based on the Xie–Beni index and an improved particle swarm optimization algorithm, a novel identification method for the Takagi–Sugeno fuzzy model is proposed in this paper. Firstly, Xie–Beni indices with a fuzzy c-means clustering algorithm are adopted to find the rule number of the Takagi–Sugeno fuzzy model. By utilizing the particle swarm optimization algorithm, the initial membership function and the consequent parameters of the fuzzy model are obtained. In addition, through an improved fuzzy c-regression model and orthogonal least-square method, the premise structure and consequent parameters can be obtained to establish the Takagi–Sugeno fuzzy model. Some well-known models are used to demonstrate that the proposed method outperforms some existing methods.