基于小波降噪和CEEMD的轴承故障诊断

针对滚动轴承故障振动信号难以提取出准确的故障特征的问题,提出了一种基于小波新阈值降噪与互补集合经验模态分解(CEEMD)的轴承故障诊断方法。该方法充分结合了以上2种方法的优点,有效地解决了故障特征提取难的问题。首先构建出新的小波阈值函数,再用此小波阈值降噪,可以有效地消除背景噪声的影响;将降噪后的故障信号用CEEMD方法进行处理,然后重构根据信号的相关系数挑选出的相关性较大的分量;最后将重构信号进行Hilbert变换包络,从包络图中提取故障特征。运用此方法对轴承进行试验分析,结果证实了该方法的有效性和实用性。

Bearing Fault Diagnosis Based on Wavelet Noise Reduction and CEEMD

Aiming at the problem that fault features are difficult to extract from the fault vibration signal of rolling bearing signal,proposed a bearing fault diagnosis method based on wavelet new threshold denoising and complementary ensemble empirical mode decomposition(CEEMD).This method fully combines the advantages of the above two methods,and effectively solves the problem of difficulty in extracting fault features.Firstly,a new wavelet threshold function was constructed.Then used the wavelet threshold to reduce noise,the influence of background noise can be effectively eliminated.The diagnosis signal completed by the noise reduction was decomposed by the CEEMD method.Corresponding components which are selected according to the correlation coefficients of the signal were reconstructed.Finally,the reconstructed signal was enveloped by the Hilbert transform and extracted fault characteristics from the envelope map.Using this method to carry out the bearing experiments analysis,the results confirm the validity and usefulness of the method.