Research On Extraction And Application Of Weak Signals Based On Wavelet Transform

In recent years, mine earthquake disasters occurs frequently and has brought great harm. Analysis and Research of mine earthquake weak signals have great significance to reduce casualties and property losses. The task of this subject is to achieve the de-noising of mine earthquake signals, get the real mine earthquake signals, lay the foundation for signal monitoring.This paper uses actual mine earthquake signals as the analysis object, lists several common used methods of weak signal processing field, analyses the advantages of wavelet transform in signal processing. This article selects the maximum reduction of real signal threshold algorithm for de-noising processing of the signal. This article introduces the four basic problems encountered in signal de-noising field, including the selection of wavelet basis function, wavelet decomposition level, threshold value and threshold function. Through a lot of matlab simulation experiment,this paper determines the wavelet basis function as sym8 and wavelet decomposition level as 3 for de-noising processing of actual mine earthquake signals. In this paper, the de-noising effect of four kinds of wavelet threshold selection rules including fixed threshold selection rule, rigrsure threshold selection rule, heursure threshold selection rule and minimax threshold selection rule is compared, it is concluded that the effect of rigrsure threshold selection rule is good. Under the same condition,this paper introduces an adaptive threshold acquisition method called Birge-Massart method,the effect of de-noising shows that Birge-Massart method is good.The innovation of this paper is to Combine Birge-Massart threshold selection rules and improved threshold function to handle mine earthquake signals, achieving the best de-noising processing mine earthquake signals. It plays a decisive role in the research and prediction of mine earthquake signals.