Learning Technique for TSK Fuzzy Model Based on Cooperative

JSK fuzzy model is decomposed into two different populations to cooperate coevolution model learning technique for its learning is the problems of multiple constraints and multiple target optimizations. All the related problems are discussed, including encode, evolution calculation, cooperation of every population and evaluation strategy of adaptive value. Fuzzy model is decomposed into two populations: one population describes fuzzy model and its rule construction, and the other population depicts fuzzy partition and membership function parameters. The technique presented has merits in little prior knowledge, rapid convergence and concise fuzzy model. Example of function approximation shows the technique´s validity.