基于经验模态分解互信息熵与同步压缩变换的微地震信号去噪方法研究  

Microseismic data denoising method based on EMD mutual information entropy and synchrosqueezing transform

作　　者：秦晅 蔡建超 刘少勇 卞爱飞 

机构地区：[1]中国地质大学(武汉)地球物理与空间信息学院地球内部多尺度成像湖北省重点实验室,湖北武汉430074

出　　处：《石油物探》2017年第5期658-666,共9页Geophysical Prospecting For Petroleum

基　　金：国家重点研发计划(2016YFC060110304);国家自然科学基金(41572116);中央高校基本科研业务费专项资金(CUG160602)联合资助~~

摘　　要：针对微地震信号具有随机性、非平稳性与时频耦合的特点以及经验模态分解(Empirical Mode Decomposition,EMD)的模态混叠问题,提出了基于经验模态分解互信息熵与同步压缩变换(Synchrosqueezing Transform,SST)的微地震信号去噪方法。首先对微地震信号进行经验模态分解,获得从高频到低频排列的固有模态函数(Intrinsic Mode Function,IMF)分量;然后求取相邻固有模态函数分量之间的互信息熵,从而辨识出高频与低频部分的分界;最后利用同步压缩变换提取高频部分的有效信号,将其与低频部分重构,实现微地震信号的有效去噪。利用不同噪声强度的理论模型和实际资料,对本文方法与直接舍弃高频成分的去噪方法进行了对比,结果表明,本文方法能够很好地去除微地震信号中的混叠噪声,并将有效信号从噪声中提取出来,提高了资料的信噪比。On the basis of the characteristics of randomness,non-stationarity, time-frequency coupling of microseismic data,and of the problem of modal aliasing in empirical mode decomposition （EMD）, this paper proposes a microseismic data denoising method based on EMD mutual information entropy and synchrosqueezing transform （SST）.First, the microseismic signal is decomposed by EMD to acquire the intrinsic mode function （IMF） sequencing from high to low frequency.Next, the mutual information entropy of adjacent IMF components is calculated to identify the boundary between the high-frequency and the low-frequency part.Finally, the effective signal of the high-frequency part is extracted by the SST and reconstructed with the low frequency part to achieve an ef fective microseismic data denoising.We applied the method to synthetic data sets with different noise intensities and to field data, and the results showed that this method can better remove the aliasing noise, extract the effective signal, and improve the SNR, compared with denoising methods that directly discard the high frequency components.

关 键 词：微地震信号 经验模态分解 同步压缩变换 互信息熵 重构 噪声压制 

分 类 号：P631[天文地球—地质矿产勘探]