• 学术论文 •

基于调控参数建模的混合模糊系统的逼近分析

1. 天津师范大学数学科学学院, 天津 300387
• 收稿日期:2015-10-21 修回日期:2016-01-19 出版日期:2017-05-25
• 通讯作者:
• 王贵君
• 作者简介:
• 董文强 男,1990年出生,硕士研究生,主要研究方向:模糊神经网络与模糊系统.E-mail:745844043@qq.com
• 基金资助:
• 国家自然科学基金 (No.61374009）

Approximation Analysis of the Hybrid Fuzzy System Based on the Modeling of a Control Parameter

DONG Wen-qiang, WANG Gui-jun

1. College of Mathematics Sciences, Tianjin Normal University, Tianjin 300387, China
• Received:2015-10-21 Revised:2016-01-19 Online:2017-05-25 Published:2017-05-25

Abstract:

A fuzzy system describes hunman language information by fuzzy rules,it is an important tool to deal with the uncertain information.Mamdani fuzzy system and Takagi-Sugeno (T-S) fuzzy system are two kinds of common systems.Firstly,the two types of systems are unified by introducing a control parameter so as to establish a generalized hybrid fuzzy system in this paper.Secondly,the hybrid fuzzy system can approximate to a continuous differentiable function with respect to infinite norm through Taylor formula with Peano remainder term of multivariate function,and certify that this system also has the second order approximation accuracy by means of Hessian matrix in Lagrange remainder term.Finally,the results show that the approximation effect of the hybrid fuzzy system based on the modeling of control parameters is better than the separate Mamdani or T-S fuzzy system by choosing sample points analysis.