基于经验模式分解和排列熵的轴承故障特征提取

Bearing Fault Feature Extraction Based on Empirical Mode Decomposition and Permutation Entropy

针对滚动轴承故障声信号故障诊断中共振解调滤波参数较难确定以及故障诊断困难的问题,提出一种基于经验模式分解和排列熵的改进滚动轴承故障诊断解调方法。该方法首先对滚动轴承声信号进行经验模式分解,将其分解为多个本征模态分量;然后计算各本征模态分量的排列熵值和相关系数,根据联合系数最大化原则对筛选出的分量进行信号重构;最后,利用快速谱峭度对重构信号进行滤波分析,将峭度值最大的频段进行平方包络提取特征频率。将该方法用于滚动轴承故障声信号的实际数据进行分析,结果表明该方法能够有效提取滚动轴承故障特征,并且相较于传统的包络解调具有更好的效果。

Aiming at the difficulty to determine the resonant demodulation filter parameters in traditional envelope demodulation for fault diagnosis of bearings,an improved fault diagnosis demodulation method for rolling bearings based on empirical mode decomposition(EMD)and permutation entropy(PE)is proposed.Firstly,the rolling bearing acoustic signal is decomposed into several eigenmode components by EMD.Then,the permutation entropy and correlation coefficient of each eigenmode component are calculated,and the filtered components are reconstructed according to the maximization principle of joint coefficient.Finally,the fast spectral kurtosis is used to filter the reconstructed signal,and the square envelope of the frequency band with the largest kurtosis value is used to extract the characteristic frequency.This method is used to analyze the actual data of rolling bearing fault acoustic signals.The results show that this method can effectively extract the fault characteristics of rolling bearings,and has a better effect than the traditional envelope demodulation.