基于盲信源分离的地球物理弱异常提取  

Detection of geophysical weak anomalies based on blind signal separation.

作　　者：左博新[1] 胡祥云[1] 

机构地区：[1]中国地质大学(武汉)计算机学院,湖北武汉430074

出　　处：《石油地球物理勘探》2014年第2期375-381,223,共7页Oil Geophysical Prospecting

基　　金：国家青年自然科学基金项目(2011093051);中央高校基本科研业务费专项基金摇篮计划项目(CUGL140401);中科院测地所重点实验室开放基金项目(SKLGED2014-1-2-E);湖北省创新群体项目(2011CDA123)联合资助

摘　　要：在地球物理勘探中探测到的深部矿体弱异常信号,由于受各种噪声和系统退化等因素的影响,通常异常体边界较为平滑和模糊,给后期解释带来一定的难度。文中提出了一种新的盲信源分离(BSS)地球物理弱异常提取算法,克服了常规地球物理空间滤波算法对于弱异常体信号无法有效提取的问题,可以在保持强异常体边界信息的同时,有效地提取弱异常体的边界。文章阐述了地球物理信号中的噪声和系统退化的理论模型,并根据推导的理论模型提出了针对地球物理数据的BSS异常提取算法。试验表明,与其他的地球物理空间滤波算法相比,本文算法具有在保持高频信息的同时,能够有效地提取低频异常信息的特点。Weak anomaly signals in geophysical deep ex-ploration are influenced by noise and systematic error degradations.This makes anomaly boundaries relatively smoothing and blurring,bring some diffculties date interpretation.This paper proposes a new algorithm for anomaly extraction based on geophysical blind signal separation(BBS),which can solves weak anomaly detection problem in general geophysical filtering algorithms.It detects highamplitude anomaly boundaries and deep weak anomaly boundaries simultaneously.This paper discusses model of noise and geophysical systematic errors degradation.The proposed geophysical anomaly detection algorithm is derived based on the model.Tests show that the proposed algorithm can extract both shallow high-frequency anomalies and deep low-frequency anomalies compared with other geophysical filtering algorithms.

关 键 词：盲信源分离 退化 噪声 边界 滤波 

分 类 号：P631[天文地球—地质矿产勘探]