基于二次EEMD的转子故障信号时频分析方法研究

HHT时频分析被广泛应用于机械故障诊断中,但其模态混叠成为应用时的瓶颈。针对此问题提出了利用二次集合经验模态分解分解(ensemble empirical mode decomposition,EEMD)来消除模态混叠的时频分析方法。该方法首先用EEMD将原信号分解成若干个本征模函数(intrinsic mode function,IMF),然后选取相关系数较大的分量重构原信号,再利用EEMD对其进行二次处理,便可获得去除模态混叠的时频分布。通过对仿真与实验转子信号分析,该方法可以有效抑制经验模式分解(empirical mode decomposition,EMD)的模态混叠现象,相比一次EEMD,二次EEMD去除模态混叠更明显,能有效应用于旋转机械故障诊断中。

StudaySonSTime-frequencySAnaylysisSMethodSof Rotor Faults SignalSBasedSonSDualSEEMD

HHT time-frequency analysis has been widely used in mechanical fault diagnosis. However, its mode mixing often brings us lot of problems in application. Aiming at this problem, a mode mixing erasing method based on the dual Ensemble Empirical Mode Decomposition was put forward. In this method, the original signal was decomposed into several Intrinsic Mode Functions (IMFs) by EEMD, Then choosing the IMFs with higher correlation coefficient, adding them together and reprocessing them with EEMD, the time-frequency distribution that was removed mode mixing was achieved. Through analysis of the simulation and experiment rotor signal, this method can effectively avoid the mode mixing phenomenon for EMD, Compared with the single EEMD, the effect of removing mode mixing by dual EEMD is more obvious. And this method can be effectively applied in rotate mechanical fault diagnosis.