基于经验模态分解与差分包络谱的齿轮故障诊断

Gear fault diagnosis based on Empirical Mode Decomposition anddifferential envelope spectrum of pure frequency modulation signal

摘要:

为了准确地进行齿轮故障诊断,结合经验模态分解与纯调频信号差分包络谱,提出了一种新的故障诊断方法。首先,运用经验模态分解对齿轮振动信号进行分解得到若干个分量；其次,根据峭度最大分量及其相邻分量的峭度值情况,合成有效分量；然后,运用经验调幅-调频分解得到纯调频信号；接着,将纯调频信号应用经验调幅-调频分解及傅里叶变换得到纯调频信号差分包络谱；最后,通过纯调频信号差分包络谱进行故障诊断。通过对齿轮断齿故障振动信号的分析,验证了方法的有效性。

Abstract:

In order to diagnose gear fault accurately, a new fault diagnosis method is proposed based on EMD and difference envelope spectrum of pure frequency modulation signal. Firstly, the empirical mode decomposition is used to decompose the gear vibration signal to obtain several components; secondly, according to the kurtosis of the maximum kurtosis value and its adjacent components, the effective components are synthesized; Next, the empirical AM-FM decomposition is used to obtain the pure FM signal; Then, the empirical AM-FM decomposition and Fourier transform are used to obtain the difference of pure FM signal Finally, fault diagnosis is carried out by the difference envelope spectrum of pure frequency modulation signal. The validity of the method is verified by processing the vibration signal of gear broken.