To suppress the random noise in microseismic signal by using empirical mode decomposition and wavelet transform

The suppression of random noise is an important step in the process of microseismic signal analysis. Nowa-days,the most of noise reduction technologies have some problems. Owing to the stochastic non-stationarity of micro-seismic signal,this paper proposed a microseismic signal denoising method combined with empirical mode decomposi-tion(EMD) and wavelet threshold denoising method to suppress random noise. The EMD can adaptively break signal down into a finite number of intrinsic mode functions (IMF) which are arranged according to the frequency from high to low order. After decomposition,the noise of microseismic signal is mainly concentrated in the higher frequency of the IMF component. Based on the principle of abrupt change of noise energy,the boundary of IMF component between high and low is found. The wavelet threshold method is used to process the high-frequency IMF. Finally,the high-fre-quency IMF components of denoised and the remaining low-frequency IMF components can be reconstructed to obtain the denoised microseismic signal. The simulation results show that the method can fully retain the transient non-statioary characteristics of the microseismic signal,and has a better denoising effect than that of the contrast method.n