基于LMD与MCKD的滚动轴承早期故障诊断方法

Early Fault Diagnosis Method for Rolling Bearings of Wind Turbines based on LMD and MCKD

当滚动轴承出现早期故障时,其故障特征信号微弱,且环境噪声较大,因此其早期故障特征一般难以提取。针对上述问题,提出基于LMD与MCKD的滚动轴承早期故障诊断方法。为了克服局部均值分解(LMD)在早期故障诊断中易受噪声影响的不足,该方法对其包含故障信号大部分能量的前4个乘积函数(product function,PF)分量进行最大相关峭度解卷积(MCKD),突出轴承信号中淹没在噪声信号中的周期脉冲成分,最后再对其进行包络解调,便可得到轴承故障特征频率,进而对滚动轴承早期微弱故障进行诊断。实验信号验证了该方法的有效性。

When the rolling bearing has early faults,the fault characteristic signal is weak and the ambient noise is large.Therefore,it is generally difficult to extract the early fault feature of the wind turbine bearing.Aiming at this problem,an early fault diagnosis method of rolling bearings based on LMD and MCKD is proposed.In order to overcome the shortcomings of local mean decomposition(LMD)in early fault diagnosis,the first four product function(PF)components,which contain most energy of fault signals,are processed with the maximum correlation kurtosis deconvolution(MCKD)to highlight the cyclic pulse component of the bearing submerged in the noise signals.Finally,the envelope demodulation is carried out for the cyclic pulse components to obtain the characteristics of bearing failure.And the early weak faults of the rolling bearing can be diagnosed.The experimental signal verify the effectiveness of the method.