基于相对小波能量与灰色相似关联度的转子系统故障诊断

Fault Diagnosis of Rotor System Based on Relative Wavelet Energy and Gray Similar Correlation Degree

针对转子系统故障时,其各频带能量分布与其故障状态之间存在映射关系,提出了基于相对小波能量与灰色相似关联度的转子系统故障诊断。采用小波分解法将轴承加速度振动信号分解为细节部分和近似部分的小波系数,这些小波系数分别对应不同频带信号,利用得到的小波系数计算相对小波能量。由于灰色相似关联度分析对小样本模式识别具有良好的分类效果,以能量分布为元素构造特征矢量,通过计算不同状态下轴承振动信号的灰色相似关联度来判断故障类型。结果表明,该方法有效应用于转子系统故障诊断。

For rootor system fault,there is a certain mapping relationship between failure state and its each frequency band energy and distribution.A rotor system fault diagnosis method based on relative wavelet energy and gray similar correlation degree is proposed.By adopting the method of wavelet decomposition to bearing vibration signal is decomposed into the details and approximation of wavelet coefficients,the wavelet coefficients corresponding to different frequency band signal,using the wavelet coefficients are obtained by calculating the relative wavelet energy.Due to the gray similar correlation degree analysis for the classification of the small sample pattern recognition has a good effect,element structure characteristic vector by energy distribution,to judge the fault types by calculating the bearing vibration signal under different status similar correlation degree.Experimental results show that the proposed can be effectively applied to rotor system fault diagnosis.