基于经验模态分解的齿轮箱故障特征提取

利用小波去噪阈值法对齿轮箱故障振动信号进行去噪,将经验模态分解(EMD)和快速傅里叶变换(FFT)相结合对齿轮箱故障进行特征提取,此方法适合于对非线性非稳态信号进行自适应的分析.利用小波阈值去噪方法对原始信号进行预处理,将去噪后的信号进行经验模态分解,得到一定数量的本征模态函数(IMF)分量,选取特定的IMF进行FFT,得到相应的功率谱,从而达到提取齿轮箱故障特征频率的目的.对齿轮箱故障信号进行分析,结果表明该方法能够有效地识别出齿轮箱故障特征频率.

Gearbox Fault Feature Extraction Based on Empirical Mode Decomposition

Wavelet threshold method was used to de-noise gearbox fault signal,the empirical mode decomposition（EMD）was associated with fast Fourier transform（FFT）to extract gearbox fault feature,and this method was suitable for analyzing non-linear and non-stationary signal adaptively.Firstly,by using wavelet threshold to preprocess the original signal,then the de-noised signal was resolved by empirical mode decomposition to get a certain number of intrinsic mode function（IMF）component and the appropriate power spectrum by selecting a specific IMF transformed by FFT,finally,thereby the aim of extracting gearbox fault characteristic frequency was attained.Through analyzing gearbox fault signal,the results showed that the method could effectively identify the gearbox fault characteristic frequency.