Application of Wavelet Soft Threshold Denoising Algorithm Based on EMD Decomposition in Vibration Signals

This paper proposes the wavelet threshold noise reduction algorithm based on Empirical Mode Decomposition (EMD) to solve the problems of centrifugal pump vibration signal complex, exist various frequency band non-single interference signal and useful signal amplitude is small, etc. The algorithm combines the adaptive characteristics of EMD decomposition and the time-frequency localization characteristics of wavelet threshold denoising algorithm. While simplifying the noise reduction process, it can effectively deal with non-single interference signals in each frequency band. In order to verify the applicability of this algorithm, it is compared with wavelet threshold noise reduction algorithm and spatial-temporal filtering analysis method. Finally, the influence of soft and hard threshold functions on the noise reduction effect is analyzed. The experimental results show that the wavelet soft threshold denoising algorithm based on EMD decomposition has better noise reduction effect when the centrifugal pump vibration signal is used as the noise reduction object.