基于局部相关谱约束的多道匹配追踪算法识别微地震信号  

Microseismic signal identification with multichannel matching pursuit based on local coherence spectrum constraint

作　　者：曹俊海 顾汉明[1,2] 尚新民[3] 

机构地区：[1]中国地质大学(武汉)地球内部多尺度成像湖北省重点实验室,湖北武汉430074 [2]中国地质大学(武汉)地球物理与空间信息学院,湖北武汉430074 [3]中国石化胜利油田分公司物探研究院,山东东营275022

出　　处：《石油地球物理勘探》2017年第4期704-714,共11页Oil Geophysical Prospecting

基　　金：国家高技术研究"863"发展计划课题"陆上非一致性时延地震;微地震油藏监测方法研究"(2011AA060303)资助

摘　　要：多道匹配追踪算法是将地震记录中的每一道信号分解成一系列相同位移参数、频率参数及相位参数但振幅不同的匹配原子的线性组合。在地震剖面中,有效信号往往具有一定的方向性。为了利用信号的方向性等特征,提出了基于局部相关谱约束的多道匹配跟踪算法。利用局部窗口内的相关谱作为权重,对每次匹配跟踪的方向进行约束,对模拟以及实际的三分量微地震信号分别采用传统的多道匹配跟踪算法和本文提出的方法进行处理。为了验证该算法的准确性及可行性,采用STA/LTA算法对其结果进行事件拾取对比分析,结果表明基于相关谱约束的多道匹配跟踪算法能显著提高微震信号的信噪比,在一定程度上降低了微震事件误判的机率,且使STA/LTA算法的峰值提高1.2倍左右,提高了微震事件检测的可能性,可较为准确地识别出微震信号。The multichannel matching pursuit(MCMP)is a lateral coherency-based approach which decomposes a trace into a series of atoms,and the atoms generated at an iteration has the same shift,frequency,scale,phase parameters,but different amplitude factor.However seismic signals are often directional.To make use of the directional feature,we propose a new algorithm,multichannel matching pursuit based on local coherence spectrum constrain.We test the proposed algorithm together with conventional MCMP on both synthetic and real microseismic data.In order to verify the accuracy and feasibility of the proposed algorithm,we pick and analyze first breaks with short time average/long time average(STA/LTA)algorithm.The processing results show that the new multichannel matching pursuit algorithm improves significantly the microseismic data signal-to-noise ratio(SNR),and reduces microseismic event false detection to a certain extent.Also the peaks of STA/LTA algorithm are improved by about 1.2times after processing.So the proposed algorithm can better identify microseismic signals than conventional MCMP.

关 键 词：多道匹配跟踪 三分量微地震记录 STA/LTA算法 信噪比 初至拾取 

分 类 号：P631[天文地球—地质矿产勘探]