基于子空间划分的模糊系统模型辨识  

Identification of Subspace-partition Based Fuzzy System Model

作　　者：白裔峰[1] 肖建[1] 

机构地区：[1]西南交通大学电气工程学院,成都610031

出　　处：《控制与决策》2006年第2期135-138,共4页Control and Decision

基　　金：教育部博士点基金项目(20040613013)

摘　　要：提出了基于子空间划分的模糊系统模型(SPFS),并给出一种针对SPSF的自适应模型辨识方法.应用遗传算法进行子空间划分方案的优化,降低了最大子空间的辨识误差,从而得到优化的模型辨识结果.理论分析和仿真计算证明了该模型的有效性.所提出的模型有助于缓解规则数爆炸问题.A subspace-partition based fuzzy system model （SPFS） and an adaptive model identification algorithm are proposed to solve the rule number＇s explosion problem. Genetic algorithm is employed to optimize subspacepartition, reduce the maximal identification error in subspaces, and partition the discourse universe on principle of consistency and completion. The relative optimum model identification result is thus achieved. The effectiveness of SPFS is proved theoretically and experimentally. The proposed SPFS model is helpful in relieving the rule number＇s explosion problem.

关 键 词：模糊系统 子空间划分 自适应辨识 

分 类 号：TP273.4[自动化与计算机技术—检测技术与自动化装置]