基于自组织神经网络的非线性系统建模

针对基于T-S模糊模型的非线性系统建模问题,提出了一种基于自组织神经网络的新方法.在T-S模糊模型的建模中,目前常用的模糊C均值聚类算法存在迭代次数多,计算耗时的缺点.首先,利用竞争学习算法对输入空间进行聚类,基于此结果,借助于模糊C均值聚类算法进一步优化聚类结果,提取T-S模糊模型的规则前件隶属函数参数.然后,采用最小二乘法求得T-S模糊模型的规则后件参数,从而建立起非线性系统的T-S模糊模型.最后,仿真结果表明,该方法可以为模糊建模提供好的模型结构,并且有较高的计算效率和精度.

Modeling of Nonlinear System Based on Self-organization Neural Map

In this paper,a novel method based on self-organization neural map is presented for fuzzy modeling of nonlinear systems based on the T-S fuzzy model.For T-S fuzzy model's system modeling,the fuzzy C-means clustering algorithm commonly used has the shortages of more iterative times and calculative time.Firstly,the input space is clustered by competitive learning algorithm,and based on the result,the clustering results are optimized by fuzzy C-means clustering algorithm,then the membership function of the antecedent fuzzy sets is obtained.Secondly,the consequent parameters of rules are calculated by the least squares estimate.Thus,the T-S fuzzy model is set up.At last,the simulation results show that the proposed method can provide a good model structure for fuzzy modeling and has high computing efficiency.