A CLUSTERING-ALGORITHM FOR FUZZY MODEL IDENTIFICATION

The fuzzy model proposed by Takagi and Sugeno can represent highly non linear systems and is widely used for the representation of fuzzy rule s. In this paper, the model is firstly modified to make its identifica tion easier. Based on the fuzzy c-partition space, four criteria are p roposed for optimization of the model parameters. Following that, a cl ustering algorithm composed of fuzzy c-linear functions clustering and like fuzzy c-means clustering is developed for minimizing the four cr iteria. An identification scheme for rule's premise and consequence pa rameters is deduced from the clustering algorithm in succession. Final ly, four examples are demonstrated to verify the effectiveness of the proposed algorithm. (C) 1998 Elsevier Science B.V. All rights reserved .