小波包能量谱在滚动轴承故障诊断中的应用

Application of Wavelet Packet Energy Spectrum in Rolling Bearing Fault Diagnosis

提出了基于小波包能量谱的滚动轴承故障诊断方法。首先,以6406轴承为例采集不同工况的振动信号,然后对试验数据进行小波包变换,振动信号被分解到独立的频段,不同频带内的信号能量变化反映了运行状态的改变,提取各频带小波包能量谱为特征向量,最后应用基于模糊聚类的故障诊断方法。经对大量实测数据的处理和分析,能够比较准确地诊断出轴承的故障。

By applying wavelet packet energy spectrum,a new method is presented for rolling bearing fault diagnosis.Firstly,the 6406 rolling bearing is taken for example,and the vibration signals are collected under different working conditions.Then,the experiment data is transformed by wavelet packet,the vibration signal is decomposed into the individual frequency bands.The variations of the signal energy in these bands reflect the different fault locations.The energy spectrum feature vectors are extracted from the individual frequency bands.Finally,the fuzzy cluster analysis method is proposed for fault diagnosis.Through processing and analyzing the massive measured data,the method is able to diagnose accurately faults of rolling bearings.