Research And Implementation Of Denoising Method For Microseismic Signals Based On Wavelet Packet Decomposition And Reconstruction

Rock burst is one of the main natural disasters in the process of mine deep mining.In order to ensure the safe production of mines,it is necessary to carry out comprehensive monitoring of rock burst.Noise reduction is the key to improve the efficiency of comprehensive monitoring.There are many noise signals in the collected signals,which are divided into random noise and non-random noise.Random noise appears randomly in the seismological records,which is irregular and difficult to be removed.Therefore,the study of random noise reduction method becomes the focus of the micro seismic signal noise reduction.Based on this,this paper mainly studies the random noise reduction method from micro-seismic signal.The traditional noise reduction method does not consider the distribution of signal and noise sufficiently,and the effect of noise reduction is not obvious.In this paper,the wavelet packet decomposition and reconstruction theory is studied and discussed deeply,and then a method of micro-seismic signal noise reduction is proposed based on the principle of wavelet packet decomposition and reconstruction.This method can suppress noise while preserving effective signals as far as possible,so as to improve the noise reduction effect of signals better.It mainly includes three aspects content as follows:(1)The noise reduction principle of traditional noise reduction methods ware summarized,and their advantages and disadvantages are compared and analyzed.In view of the related theories involved in the denoising method in this paper,the article gives a detailed description,which provides a theoretical basis for the later denoising method and the orderly conduct of the experiment.(2)A noise reduction method for microseismic signals is studied and proposed..the basic idea of this method is summarized as follows:according to the sampling frequency of the picker,the N layer wavelet packet decomposition of the micro-seismic signal is realized,and the sub-band of 2N is obtained.Calculating correlation coefficient between each sub-band and the original signal,the sub-band with a large correlation coefficient is not processed,and sub-band with a medium correlation coefficient takes threshold noise reduction,the remaining sub-band is directly excluded,and finally the signal after be processed are reconstructed to get the final signal after noise reduction.Selection and quantification of related parameters about noise reduction is given.In order to make the effect of noise reduction more reasonable and effective,and different parameters ware compared and analyzed in detail.(3)A microseismic noise reduction system based on Matlab/GUI development platform is researched and developed,and noise reduction theory and process are applied on the system.Through simulation analysis and example analysis,compared with the traditional Fourier transform noise reduction method,median filter method and empirical mode decomposition method,the noise reduction method proposed in this paper is compared with the traditional Fourier transform noise reduction method,median filter method and empirical mode decomposition method.Observing the noise reduction effect of different methods to verificate effectiveness of this method.Signal noise ratio,root mean square error and percentage of energy are introduced as evaluation criteria of noise reduction effect.The simulation results show that the signal after noise reduction better retained the signal characteristics,it has higher signal to noise ratio,lower root mean square error and higher energy percentage.The real micro-seismic signal is processed to engineer by this method,and also achieved good results.A comprehensive analysis shows that the method can be applied to the noise reduction processing of micro-seismic signals.