基于小波和EMD的时频矩阵DEM降噪方法

Time Frequency Matrix DEM De-noising Method Based on Wavelet and EMD

针对强噪声背景下振动信号容易被非平稳强噪声干扰问题,提出了一种小波和EMD结合的时频矩阵DEM降噪方法,该方法拟解决小波变换对信号局部自适应性差的问题,同时增强EMD本征模态函数的选取和重构。通过该方法对仿真信号和实测振动信号降噪,结果表明:该方法能够有效提取强噪声背景下有效的振动信号,提高信噪比。这一方法为旋转机械在强噪声干扰运行环境下提取有效故障信息提供支持。

For the non-stationary interference problem of vibration signal in the background of strong noise, the time frequency matrix DEM de-noising method combining wavelet and EMD was proposed to solve the poor local adaptability of wavelet transformation to signal and reinforce the selection and reconstruction of the EMD IMFs. The method was employed to de-noise simulated and actual vibration signals, and the findings indicated that the method could be effective in extracting valid vibration signal in the background of strong noise and increasing the signal-to-noise ratio. The method offers support for the extraction of valid fault information from rotating machines in the operating environment of strong noise disturbance.