基于EMD的时频熵在齿轮故障诊断中的应用

提出了一种基于EMD(EmpiricalModeDecomposition)方法的时频熵齿轮故障诊断方法。首先利用EMD方法分解齿轮振动信号,然后将得到的内禀模态分量进行Hilbert变换,以得到振动信号的时频分布,将信息熵理论引入时频分布,定量描述时频平面上不同时频段的能量分布,各时频段能量分布的均匀性可以反应齿轮的运行状态的差别,从而可以通过时频熵的大小判断齿轮的工作状态和故障类型。实验证明该方法能有效的判断齿轮故障特征,为齿轮故障诊断提供了新的思路。

APPLICATION OF TIME-FREQUENCY ENTROPY TO GEAR FAULT DIAGNOSIS BASED ON EMD

A gear fault diagnosis approach based on EMD is proposed. In the method the time-frequency entropy is introduced. Firstly the gear vibration signal is decomposed by EMD. Then the resulting IMF components are transformed by Hilbert transformation to find the time-frequency distribution of the vibration signal. Finally the information entropy theory is introduced to describe quantitatively the energy distribution of different time-frequency ranges in the time-frequency plane. The uniformity of energy distribution of each time-frequency range can be used to reflect the difference of the working state of gears so that the fault type of gears can be diagnosed. It is shown by the experiment results that this method can be applied to identify the gear fault effectively and offers a new way for the gear fault diagnosis.