摘要
基于故障轴承的特征提取,提出了将小波包分析与神经网络结合的滚动轴承故障诊断方法。对滚动轴承信号进行3层小波包分解,构造小波包特征向量作为故障样本,用训练好的BP神经网络进行故障诊断,试验结果表明,该方法能够有效地诊断出滚动轴承的故障类型。
The method of wavelet packet transform and neural network is presented to diagnose rolling bearings faults based on feature extracting of fault bearing.Three-layer wavelet packet is adopted to decompose the signal of rolling bearings,and wavelet packet energy eigenvector is constructed as fault samples,then the trained three-layer BP neural network is used to diagnose fault.The practical example shows that the method is able to diagnose the kinds of rolling bearings faults.
出处
《轴承》
北大核心
2008年第4期46-48,共3页
Bearing
关键词
滚动轴承
故障诊断
小波包特征向量
神经网络
rolling bearing
fault diagnosis
wavelet packet eigenvector
neural network
