基于加强谱峭度的航空发动机齿轮毂故障诊断

An Enhanced Spectral Kurtosis Methodforaero-Engine Gear Hub Fault Diagnosis

关键词:  航空发动机  故障诊断  AR模型  小波包  谱峭度

Key words:  Aero-engine  Fault diagnosis  AR model  WPT  Spectral kurtosis

摘要:

针对航空发动机减速器一级齿轮毂故障诊断问题，提出了一种基于小波包和CHI-LMD(三次Hermite插值-局部均值分解)的加强谱峭度的故障诊断方法。在用AR(自回归)参数模型对原始信号进行降噪后，首先采用小波包对信号进行分解，并结合谱峭度找出特征频带，继而用CHI-LMD对特征频带进行再分解获得若干PF分量，最后对获得的PF分量计算谱峭度作为故障识别参数。利用此方法对10组待识别信号的诊断结果表明，该方法能有效识别减速器一级齿轮毂故障，在不拆卸发动机的情况下实现了对目标的诊断。

Abstract:

As for the issues of fault diagnosis for reducer gear hub of the aero-engine，an enhanced spectral kurtosis method based on the wavelet packet transform(WPT)and cubic Hermite interpolation-local mean decomposition(CHI-LMD)was presented. Under the noise reduction of autoregressive(AR)model，firstly the signal was decomposed by WPT and the feature band with spectral kurtosis was obtained. Then，feature band was decomposed by CHI-LMD to get product functions(PFs)and finally took spectral kurtosis of PF as the final diagnostic parameters. The diagnosis adopting this method on 10 groups of undetermined signal indicated that the proposed method could determine the fault of reducer gear hub efficiently，which realized the fault diagnosis of target without dismantling aero-engine.