基于经验模态分解的滚动轴承故障诊断方法

提出了一种基于经验模态分解的滚动轴承故障诊断方法，并定义了能量熵的概念。从不同状态的滚动轴承振动信号的能量熵值中发现，当滚动轴承发生故障时，各频带的能量会发生变化。为了进一步对滚动轴承的状态和故障类型进行分类，再从若干个包含主要故障信息的IMF分量中提取能量特征参数作为神经网络的输入参数来识别滚动轴承的故障类型。对滚动轴承的正常状态、内圈故障和外圈故障振动信号的分析结果表明，以经验模态分解为预处理器提取各频带能量作为特征参数的神经网络诊断方法比以小波包分析为预处理器的神经网络诊断方法有更高的故障识别率，可以准确、有效地识别滚动轴承的工作状态和故障类别。

Roller Bearing Fault Diagnosis Method Based on EMD

Roller bearing fault diagnosis method based on Empirical Mode Decomposition (EMD) was put forward and the concept of EMD energy entropy was introduced. The analysis results from energy entropy of different vibration signals show that the energy of acceleration vibration signal will vary in different frequencies bands when bearing faults occured. To identify roller bearing fault patterns, energy feature parameters extracted from a number of IMFs which contained main fault informations can be served as input parameters of the neural network to identify fault patterns. The analysis results from roller bearing signals with inner-race and out-race faults show that the diagnosis approach of neural network based on EMD extracting energy of different frequencies bands as features is superior to that based on wavelet packet decomposition and reconstruction and would identify roller bearing fault patterns accurately and effectively.