自相关去噪和经验模态分解的轴承故障分析

为提取滚动轴承故障诊断中的信号,给出一种分离故障信号的新方法。该方法先对原始信号进行自相关去噪,再对去噪信号进行经验模态分解(EMD),得到各个本征模态函数(IMF),计算各个本征模态与去噪信号的相关系数。选择相关系数绝对值较大的本征模态进行重构,再对其去噪。最后对IMF重构信号进行包络谱分析,可以清晰地看到故障特征频率。通过仿真分析和两种不同轴承试验结果表明,所提出的方法可以有效地抑制噪声,并能得到反映实际故障信息的信号。

Bearing Fault Analysis of Autocorrelation Denoising and Empirical Mode Decomposition

In order to extract signals from fault diagnosis of rolling bearings,a new method for separating fault signals is presented.First,the original signal goes through autocorrelation noise reduction.Empirical mode decomposition is used to denoise signals,and the eigenmode function is obtained.Second,the correlation coefficient of each intrinsic mode and the denoising signal are calculated.The absolute value of correlation coefficient which is larger than other absolute values is reconstructed,and then we denoise it.Finally,the characteristic frequency of fault can be shown by envelope spectrum analysis of IMF reconstructed signal.The analyzed results of simulation analysis and two different bearings validate that the proposed method can effectively suppress the noise and get the signal which reflects the actual fault information.