优化参数VMD和MED在列车齿轮箱滚动轴承故障诊断中的应用

Application of Optimization Parameters VMD and MED in Fault Diagnosis of Train Gearbox Rolling Bearings

针对强噪声情况下列车齿轮箱滚动轴承早期故障特征提取困难的问题,提出基于最小熵解卷积(minimum entropy deconvolution,MED)与参数优化变分模态分解(variational mode decomposition,VMD)相结合的故障诊断方法。首先利用MED对轴承振动信号进行降噪;其次,采用离散差分进化算法(discrete differential evolution algorithm,DDE)对VMD的参数进行优化搜索,并利用优化参数的变分模态分解算法对降噪后的故障信号进行处理,得到一系列本征模态函数;最后,选择最佳的本征模态函数进行包络分析,从而提取出故障特征。试验结果表明,该方法能有效提取列车齿轮箱滚动轴承故障特征,可用于轴承故障诊断。

Aiming at the problem of feature extraction of train gearbox rolling bearing’s incipient fault in the case of strong noise,a method of fault diagnosis based on minimum entropy deconvolution(MED) and parameter optimized variational mode decomposition(VMD) was proposed.Firstly,the bearing vibration signal was denoised by using MED.Then,the VMD parameters were optimized by discrete differential evolution algorithm(DDE),and the denoising signal was processed by VMD using the optimum parameters obtained by searching,a series of intrinsic mode functions were obtained.Finally,the optimal intrinsic mode function(IMF)was selected for envelopment analysis and getting the fault frequency.The experimental results showed that the proposed method could effectively extract the fault features of train gearbox rolling bearing and could be used to rolling bearing faulf diagnosis.