基于复Morlet小波和系数相关的齿轮故障特征提取

On Fault Feature Extraction of a Gear by Complex Morlet Wavelet Transform and Coefficient Correlation

针对大型机械测取的振动信号信噪比低,故障特征不明显,故障定位难度大,提出了基于复Morlet小波和系数相关的齿轮故障特征提取方法。该方法利用了复Morlet小波的幅值、相位组合信息对信号突变点具有更好的敏感特性和小波系数相关降噪特性,对被测信号进行复Morlet小波变换,再分别将小波系数的实部和虚部进行自相关处理,并将相关后系数的幅值和相位进行组合。该方法在对齿轮传动弱故障信号特征提取的试验结果表明,该方法与直接的复Morlet小波变换相比,能够有效去除噪声,更好地突出故障特征,对故障特征点进行更精确地定位。

When a large machine is working,the signal-to-noise ratio（SNR） of its vibration signal is very low,the fault feature is not distinct.A new method based on complex Morlet wavelet transform and coefficient correlation was introduced to extract the fault feature of a gear.It makes use of the compounding information composed of the magnitudes and phases of complex wavelet transform and coefficient correlation filter characteristic.Measured signal is first processed by complex Morlet wavelet transform.Then wavelet coefficient is correlated and the magnitudes and phases of the correlation coefficient are compounded.The fault signal of the gear was analyzed by this method.The results show that this method can considerably suppress the noise and position the fault feature point more accurately.