基于小波包变换的滚动轴承故障诊断方法的研究

目前基于小波分析的滚动轴承故障诊断方法的研究已经很多，但是这些方法对于强噪声背景下的故障信号特征提取效果并不理想。为此，提出了适用于强噪声背景的自相关及互相关小波包消噪滚动轴承故障诊断方法。该方法首次将相关分析和小波包分解结合：对被测信号进行自相关或互相关处理，之后进行小波包阈值消噪处理，对消噪最大能量系数进行自相关或互相关处理，最后对能量序列进行FFT计算。仿真结果表明，该方法极大地增强了对滚动轴承故障诊断的能力，在强噪声背景下有效地提取出滚动轴承的故障频率。

Research on Fault Diagnosis of Rolling Elements Bearing Based on Wavelet Packets Transform

There has been a lot of research for diagnosing rolling element bearing faults using wavelet analysis,but these methods are usually not very ideal for picking up fault signal characteristis under condition of strong noise.In the paper,a method in which correlation analysis is combined with wavelet packets transform(WPT)for the purpose of denoising is proposed.According to the method,the procedure is as follows:computing auto-correlation or cross-correlation of the measured signals,de-noising by thresholding,computing auto-correlation or cross-correlation of maximal energy coefficients and then carrying out FFT of energy sequence.The simulation results reveal that the method boosts up the capability of feature extraction and fault diagnosis for rolling bearings.