参数优化VMD结合1.5维谱的滚动轴承复合故障特征分离方法

Compound fault features separation method of rolling bearing based on parameter optimization VMD and 1.5 dimension spectrum

针对滚动轴承复合故障分离的问题,基于相关峭度具有突出故障冲击周期性的特点和1.5维谱抑制高斯白噪声、剔除非耦合谐波分量的优点,提出了参数优化VMD结合1.5维谱的滚动轴承复合故障特征分离的方法。首先以轴承不同故障特征频率对应的周期计算得到的最大相关峭度为评价指标,通过相应的相关峭度图来实现VMD中参数选择以及最优分量的选取;然后对最优分量进行包络处理,并为减少冗余成分和噪声干扰,选择1.5维谱来对包络信号做进一步分析,以此来实现滚动轴承复合故障的有效分离。通过对轴承复合故障仿真及实验信号的分析证明了该方法的有效性。

Aiming at the problem of compound fault separation of rolling bearings,based on the correlation kurtosis has the characteristic of stressing the periodicity of fault shock and 1.5 dimension spectrum has the advantage of suppressing Gauss white noise and eliminating uncoupled harmonic components,a method of feature separation for compound fault of rolling bearings based on parameter optimization variational mode decomposition(VMD)and 1.5 dimension spectrum was proposed.Firstly,the maximum correlation kurtosis from the period corresponding to different fault characteristic frequencies of bearing was taken as the evaluation index,and the parameter and the optimal component selection in VMD were realized by the correlation kurtosis diagram.Then the optimal component was enveloped.In order to reduce redundant components and noise interference,1.5 dimension spectrum was selected for further analysis,and to achieve effective separation of compound fault of rolling bearings as well.The validity of this method was proved by the simulative and measured signals of bearing compound faults.