摘要针对齿轮故障振动信号的非平稳调制特性以及传统共振解调方法不易确定滤波器参数的缺点,提出了一种基于局部均值分解(Local mean decomposition,LMD)时频分析的谱峭度 (Spectrum Kurtosis,SK)分析方法,并将其应用于齿轮故障诊断。该方法首先利用LMD对齿轮故障振动信号进行分析得到时频分布,然后将时频分布按照不同的尺度分成若干不同的频段,计算每一频段内信号的谱峭度值,并得到相应的峭度图,再根据峭度最大原则选取滤波频段,对滤波后的信号进行包络分析以获得齿轮振动信号的故障信息。利用该方法分别对仿真信号以及齿轮故障振动信号进行了分析,结果表明,基于LMD的谱峭度分析方法能够有效地提取齿轮故障振动信号特征。
Abstract:Aiming at the nonstationary modulation characteristic of the gear fault vibration signal and the drawback that the filter parameters can’t be defined effectively in traditional resonance-demodulation method, the spectrum kurtosis approach based on local mean decomposition is introduced into gear fault diagnosis. In this approach, firstly LMD method is used to analyzing the gear vibration signal and the time-frequency distribution can be obtained; then, the time-frequency distribution is decomposed into some frequency band by using different scale and the kurtogram can be obtained with calculating the spectrum kurtosis of each frequency band; finally the filtering band can be selected by the max kurtosis and the fault characteristics frequency of the signal can be extracted by envelope analysis for filtered signal. The proposed approach has been applied to simulated signal and gears fault diagnosis. The analysis results show that the spectrum kurtosis approach based on LMD can extract the characteristics of gear fault vibration signals efficiently.
程军圣;杨怡;杨宇. 基于LMD的谱峭度方法在齿轮故障诊断中的应用 [J]. , 2012, 31(18): 20-23.
Cheng Jun-sheng;Yang Yi;Yang Yu. Application of the spectrum kurtosis approach based on local mean decomposition to gear fault diagnosis. , 2012, 31(18): 20-23.