基于Hilbert谱奇异值的滚动轴承故障诊断

提出一种基于Hilbert谱奇异值的故障特征提取方法，将其与支持向量机结合应用于轴承故障诊断。利用小波阈值降噪的方法对拾取的轴承故障振动信号进行滤波降噪，然后利用经验模式分解将降噪信号分解为若干个IMF分量之和，对每个IMF分量进行Hilbert变换得到振动信号的Hilbert谱，对Hilbert谱进行奇异值分解得到反映轴承状态特征的奇异值序列，再利用奇异值作为特征向量，应用支持向量机进行轴承故障诊断，并对不同核函数的诊断结果进行了分析比较。对正常轴承、内圈故障、外圈故障、滚动体故障的实际信号的诊断验证了该方法可的有效性。

Rolling Bearing Fault Diagnosis Based on Hilbert Spectrum Singular Value Decomposition

A feature extraction method based on Hilbert spectrum and singular value decomposition in combination with the support vector machine （SVM） is proposed and applied to bearing fault diagnosis. Firstly, the wavelet threshold de-noising method is used to extract the fault information from vibration signals. Then, the de-noised signal is decomposed into several IMF components through the empirical mode decomposition method, Hilbert transform is applied to each IMF component and the Hilbert spectrum of the signals is obtained. The singular decomposition value method is used to the Hilbert spectrum to extract the time-frequency feature of the faulted bearing. Finally, the singular values are used as the feature vector to identify the different faults by using the SVM method with different kernel functions, and the diagnosis result is analyzed and compared. The results of the diagnosis of the bearing with the inner-race fault, ball fault and outer-race fault verify the effectiveness of the proposed method.