A novel T–S fuzzy systems identification with block structured sparse representation

In this paper, we propose a fuzzy partition based T–S fuzzy systems identification method with block structured sparse representation. Firstly, a novel fuzzy partition method is developed to learn fuzzy rule dictionaries by taking advantage of the geometrical structure of input variables and the functional relationship between input and output variables. Then, we explicitly focus on the block structured information existing in T–S fuzzy models and cast the systems identification problem into an optimization problem with structured sparse representation. In such a way, accurate description of T–S fuzzy model is established with far fewer numbers of fuzzy rules by selecting the important fuzzy rules and eliminating the redundant ones in the process of block structured sparse regression. Several numerical experiments on well-known benchmark data sets are carried out to illustrate the effectiveness of the proposed method.