最优参数MCKD与ELMD在轴承复合故障诊断中的应用研究

Application of OPMCKD and ELMD in bearing compound fault diagnosis

机械设备中滚动轴承复合故障的情况普遍存在。针对多种故障难分离和提取的问题,提出了基于最优参数最大相关峭度解卷积(Optimal Parameter Maxim Correlated Kurtosis Deconvolution,OPMCKD)与总体局部均值分解方法(Ensemble Local Mean Decomposition, ELMD)相结合的轴承复合故障诊断方法;首先利用排列熵值、包络谱稀疏度分别筛选MCKD中的最优滤波器长度L与冲击周期T,提取滚动轴承主故障;然后通过ELMD方法将非平稳信号分解为若干个分量,筛去主故障信息后,再次利用最优参数MCKD进行次故障诊断。通过对轴承信号的分析,验证了该方法能有效分离复合故障信号,具有一定的实用性。

Compound faults of rolling bearings exist ubiquitously in mechanical equipment. Aiming at problems of multi-fault being difficult to separate and extract, a bearing compound fault diagnosis method based on the optimal parameter maximum correlated kurtosis deconvolution(OPMCKD) and the ensemble local mean decomposition(ELMD) was proposed. Firstly, the optimal filter length L and the shock period T in OPMCKD were screened, respectively using the permutation entropy values and the envelope spectrum sparsity to extract the main fault in rolling bearing. Then ELMD method was used to decompose nonstationary fault vibration signals into several components. After screening out the main fault information, OPMCKD was used to perform the secondary fault diagnosis. Through analyzing actual bearings’ fault vibration signals, it was shown that OPMCKD and ELMD can be used to effectively separate bearing compound fault signals, and have a certain practicability.