Application of RBF neural network based on ENN2 clustering in Fault Diagnosis

被引：2

作者：

Wen, Tianzhu ^{[1
]}

Xu, Aiqiang ^{[1
]}

Liu, Chunxia ^{[1
]}

Li, Nan ^{[2
]}

机构：

[1] Naval Aeronaut & Astronaut, Yantai, Peoples R China

[2] PLA 91395, Beijing, Peoples R China

来源：

2014 SIXTH INTERNATIONAL CONFERENCE ON INTELLIGENT HUMAN-MACHINE SYSTEMS AND CYBERNETICS (IHMSC), VOL 2 | 2014年

关键词：

fault diagnosis; RBF neural network; ENN2; clustering; extension theory; unsupervised learning;

D O I：

10.1109/IHMSC.2014.120

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Radial basis function (RBF) neural network is widely used in engineering with its powerful advantage in solving nonlinear problems. But the number of hidden layer as well as the center and standard deviation of radial basis function are difficult to get, so RBF neural network based on ENN2 is proposed to solve the fault diagnosis problem. Firstly, the structure of RBF neural network is introduced; afterwards, the learning algorithm of RBF neural network is analyzed, the center and standard deviation of RBF in hidden layer are obtained by clustering method of extension neural network type 2(ENN2), meanwhile the weight matrix between hidden layer and output layer are calculated by generalized inverse method. Ultimately, the method is used to solve fault diagnosis problem, the results show that it has the advantages of simple structure, fast learning speed and high diagnostic accuracy.

引用

页码：71 / 74

页数：4

共 13 条

[1]

Mohamed S, 2006, IEEE IJCNN, P3942

[2]

Wang M. H., 2003, IEEE T NEURAL NETWOR, P339

[3] Extension neural network-type 2 and its applications [J].