摘要
提出了一种基于蚁群聚类的模糊神经网络算法,神经网络采用RBF网络结点结构,聚类采用二级结构蚁群聚类算法作为一级聚类而模糊C-均值聚类(FCM)用于二级聚类。将上述聚类方法用于模糊神经网络构建中,仿真结果表明具有并行实时性、聚类能力强的特点。
Fuzzy neural network based on fuzzy C-means clustering algorithm (FCM) is applied widely to the design of modeling and controlling. This paper proposes a fuzzy neural network based on two-stage clustering algorithm in order to determine the fuzzy rules and the number of rules. The neural network based on RBF is constructed using the relationship between the neural network and RBF. The way provides the solve of optimal parameter which consist of parallel just-time, the strong ability of clustering.
出处
《华东理工大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2005年第2期215-218,共4页
Journal of East China University of Science and Technology
基金
国家863资助课题(2002AA412120)
