基于模糊聚类的非线性系统模糊辨识算法

Fuzzy Identification Algorithm for Nonlinear Systems Based on Fuzzy Clustering

针对模糊聚类算法中数据和运算耗时很长,不适于在线建模与控制的问题,基于模糊聚类型隶属函数和EUM方法,提出了一种新的模糊辨识算法。该方法省去了求解聚类中心的迭代过程,计算时间显著减少。采用该方法对Box-Jenkins煤气炉数据和Mackey-Glass混沌时间序列进行了仿真,结果证明了该方法的有效性。

To the problems of time-consuming and unsuitability for on-line modeling and control, based on the fuzzy clustering membership function and equalized universe method, a new fuzzy identification algorithm is proposed. The iteration process for researching cluster centers is not necessary, and computation time required to partition a data set into C classes is quitely reduced. The simulation results on gas stove data and Mackey-Glass time series show the effectiveness of the method.