AN EXTENSION-PRINCIPLE-BASED FUZZY MODEL AND ITS IDENTIFICATION ALGORITHM
一种基于扩张原理的模糊模型及其辨识方法

An extended TSK model--ETSK model, based on LR type fuzzy numbers and their extended operations, is proposed in this paper. The input-output analytic expression is proved and a parameter identification algorithm is also constructed to identify the areas and gravity centers of the membership functions of LR type fuzzy numbers. A kind of variable weight TSK model which is equivalent to the ETSK model is also deduced. Simulation shows that the ETSK model can give out more accurate long-range predictions than an LM model and a TSK model.