摘要
对轴承故障信号进行3层小波包分解,重构第3层所有节点,提取重构信号频谱的峰值作为故障特征点并构成特征空间,计算特征空间的平均欧氏距离,平均欧氏距离最小时对应的节点即为最优小波包节点,重构最优节点得到最优重构信号并从中提取特征点构成最优特征空间,最后,对最优特征空间进行K均值聚类。对4种转速下轴承的4种状态进行特征提取与模式识别试验,结果表明,运用该方法能有效提取轴承故障的特征,并使故障特征空间具有最低的类内离散度,获得了较高的模式识别准确率。
First of all,the bearing fault signals were decomposed into three layers wavelet coeffi-cients by which the bearing fault signals were re-constructed.The peak values extracted from the reconstructing signal spectrum constructed a fea-ture space.Then,the minimum average Euclidean distance calculated from the feature space indicated the optimal wavelet packet node.The optimal fea-ture space could be constructed by the feature points extracted from the signals reconstructed by the optimal wavelet packet nodes.Finally,the opti-mal feature space was used for the K means clus-tering.The feature extraction and pattern recogni-tion test of the four kinds of bearing conditions un-der four kinds of rotation speeds was detailed.The test results show this method,which can extract the bearing fault feature efficiently and make the fault feature space have the lowest within class scatter,wons a high pattern recognition accuracy.
出处
《机械与电子》
2014年第5期12-16,共5页
Machinery & Electronics
基金
云南省自然科学基金重点项目(2010CD030)
关键词
轴承故障
特征提取
小波包分解
最优节点
K均值聚类
模式识别
bearing fault
feature extraction
wavelet packet decomposition
optimal node
K means clustering
pattern recognition