典型非线性模型的模糊辨识逼近精度分析

Approximation Precision Analysis of Typical Nonlinear Model Identification Using Fuzzy System

人们根据非线性系统的复杂特性归结了几种具有代表性的非线性模型。而模糊辨识方法是辨识非线性系统的有力工具,本文采用T-S模糊模型对三种常见的非线性模型:Hammerstein模型,Wiener模型和双线性模型进行逼近,并根据仿真数据研究不同的非线性结构对模糊模型逼近精度的影响。仿真实例是在训练和检验数据组数、模型阶数相同的情况下,采用三角形隶属函数,聚类型隶属函数和高斯型隶属函数分别对这三种非线性模型进行逼近能力的研究。

Based on the different nonlinear peculiarities, several nonlinear models have been defined. The fuzzy identification method is regarded as a powerful instrument in nonlinear system identification. In this paper, we studied about three nonlinear models which are Hammerstein model, Wiener model and singular model. These models will be identified by using the T-S fuzzy model, and the influences in fuzzy model accuracy which are brown by the different nonlinear structures will be seen from the simulation data. Based on the same rank and the same queue of training and test input, the triangle membership function, clustering membership function and gauss membership function were applied to prove. The simulation shows the approximation precision of the above nonlinear models.