摘要
介绍了模糊聚类方法的原理及FCM算法,并将它应用于少故障类别的轴承故障诊断,取得了良好的效果,与神经网络方法相比,模糊聚类方法的实现只需少量样本,克服了神经网络方法需要大量样本来实现的瓶颈．
This paper shows the principle of fuzzy clustering and arithmetic of FCM. When fuzzy clustering is applied to fault diagnosis of bearing, it can get an ideal result. The method of fuzzy clustering is more practical than neural network, because it only needs a few samples while neural network needs a great deal of samples.
出处
《江西理工大学学报》
CAS
2006年第4期15-17,共3页
Journal of Jiangxi University of Science and Technology
关键词
模糊聚类
故障诊断
fuzzy clustering
fault diagnosis
