ETSK模糊模型辨识及模糊控制算法

本文提出一种基于扩张原理的ＥＴＳＫ（ＥｘｔｅｎｄｅｄＴＳＫ）模型，导出了该模型的输入输出解析式，给出了辨识这种模型的方法。本文还导出了ＥＴＳＫ模型的一种等价形式——变权ＴＳＫ模型，从而将ＥＴＳＫ模型规则后件中的模糊数及其扩展运算转化为普通数的运算，使基于ＥＴＳＫ模型的模糊控制算法ＭＢＦＣ（Ｍｏｄｅｌ－ＢａｓｅｄＦｕｚｚｙＣｏｎｔｒｏｌ）易于实现。仿真辨识结果表明，ＥＴＳＫ模型的辨识效果和预报精度优于ＴＳＫ和ＬＭ模型；ＭＢＦＣ算法的控制效果优于通常模型ＰＩ控制算法

The Identification of ETSK Fuzzy Model and a Kind of Fuzzy Control Algorithm

This paper presents an extension principle based fuzzy model named ETSK model(Extended Takagi Sugeno Kang model). Under some giver conditons,the analytic expression of ETSK model is derived and an algorithm to identify such model is proposed. Though ETSK model has strong interpreting ability,the fuzzy numbers and extended operations in the consequent of the rules make it very difficult to design a fuzzy controller. To slove this problem,a variable weights TSK model which is equivalent to ETSK model is deduced. Then an ETSK model based fuzzy control algorithm named MBFC algorithm is presented,in which the fuzzy control rules are designed according to rules of the variable weights TSK model. It is show that ETSK model can give out more accurate long range predictions and MBFC algorithm can achive better control performance.