基于小波包联合AR功率谱理论的滚动轴承故障诊断

提出了小波包联合自回归功率谱理论的故障诊断方法.对采集的轴承振动信号采用高、低正交共轭镜面滤波器组,将信号划分到不同频道上.滤波器每作用一次,数据点减半,采样的时间增至两倍.选取轴承缺陷所在频段的数据插零,将其他频带补零重构提高缺陷信号的时频分辨率;然后通过AR功率谱分析轴承运行状态,诊断出轴承对应的故障.对207滚动轴承的早期缺陷作了实际诊断,诊断结果与实际较为符合.证明该方法是一种有效的弱信号缺陷提取与诊断方法.

Fault diagnosis of rolling bearing based on wavelet packet combined autoregressive power spectrum theory

A fault diagnosis method is proposed based on wavelet packet combined autoregressive power spectrum. In this method, high and low frequency quadrature conjugate mirror filter banks is used to project rolling bearing sampling signal onto different frequency bands; the number of data is decreased to half and sampling time interval is doubled by every filtering; to get higher time and frequency resolution, zero is interpolated into the frequency bands of the defects , zero is added to other bands, and then the signal is reconstructed; the work state of the bearing can be analyzed according to autoregressive power spectrum of the bearing. The diagnosis result of early defect of 207 rolling bearing tallies with the actual situation, which testifies that the method is effective to the extraction and diagnosis of the weak signal of defects.