Research on fuzzy modeling based on modified fuzzy clustering and its application

A new method of fuzzy modeling based on T-S fuzzy model is presented. In this approach, a modified fuzzy clustering algorithm is combined with the least square approximation method to identify the structure and parameters of the fuzzy model from a set of data. An L-M technique is used to optimize initial fuzzy model. The proposed method is applied to coke-oven temperature system, and the simulation results demonstrate its effectiveness.