摘要
提出小波去噪和EMD相结合的齿轮箱故障诊断的新方法。该方法首先对原始信号进行小波阈值去噪,将去噪信号利用EMD方法分解为多个IMF分量,计算各IMF分量和原信号的互相关系数,选择互相关系数较大的IMF分量进行Hilbert包络谱分析,提取故障频率。以互相关准则提取IMF分量避免了IMF分量选择的盲目性。对实测齿轮箱故障信号进行了分析,结果表明该方法能够有效地识别齿轮箱故障频率。
A new method combining wavelet denoising and EMD （empirical mode decomposition） for gearbox fault diagnosis is proposed. The noise of gearbox fault signal will be removed by wavelet thresholding denoising with this method, the denoised signal will be decomposed into several IMFs （intrinsic mode functions） by EMD, the cross-correlation coefficients between each IMF and original signal are calculated, the IMFs corresponding to bigger coefficients are analyzed with Hilbert envelope spectrum, and fault frequency will be got form envelope spectrum. Choosing IMFs with cross- correlation criterion avoids blindness of IMF components selection. Factual gearbox fault signal is analyzed, and result shows that fault fre^uencv of ~earbox can be identified effeetivelv.
出处
《煤矿机械》
北大核心
2012年第4期278-280,共3页
Coal Mine Machinery
关键词
小波去噪
EMD
Hilbert包络谱:互相关
wavelet denoisin
empirical mode decomposition
Hilbert envelope spectrum
cross--correlation
