基于独立分量分析的油气富集区地震识别.docxVIP

基于独立分量分析的油气富集区地震识别 王任一，易平，马勇新，温伟明 （中海石油湛江分公司勘探开发部 广东湛江 524057） 摘要：地震记录是地下各种地质因素和噪音的综合响应。地下地质体可等效分成地质体几何 形态、地质体骨架属性、其中流体属性和噪音四个相互独立“基本结构因子”，地震观测信 号可等效看成是地下地质体四个“基本结构因子”作为信号源的线性混叠信号，利用ICA 技术可从混叠信号中提取出这些相互独立信号源，从而可提取出隐含在地震观测信号中的含 油气性（流体属性）和沉积相（骨架属性）等有用的信息。利用ICA从地震属性中分离出相互独 立信号源的地质解释，主要依靠对工区现有的地质认识成果来赋予它实际的地质含义，这具 有一定的主观随意性，可通过信息嫡等其他数学方法减小这种解释的主观性。 关键词：油气富集区；沉积相；独立分量分析；地震属性；混叠信号；信息爛 Seismic identification in oil-gas accumuI at i on area us i ng independent component analysis technique Wang Renyi, Yi ping, Ma Yongxin, Wen Weiming （ZhanJiang Branch Company of Petroleum Exploration and Development, CNOOC,Zhanjiang, Guangdong, 524057） Abstract: Seismic recording is synthetic response of underground various geologic factors and noises. Subsurface geologic bodies are equivalent to four independent elementary structure factors of geologic body geometry, geologic body matrix properties, liquid properties and noises. Seismic observation signal is equivalent to linearity aliasing signal of signal source mentioned in the above elementary structure factors. The independent signal source can be extracted from aliasing signal by the use of ICA? Accordingly some available information can be extracted such as petroliferous and sedimentary microfacies information which connotatively exists in seismic observation signal. Geologic explanation of independent signal source separated from seismic attribute by the use of ICA is provided with subjectivity as a result of actual geologic significance given by existing geologic research results, which can reduce explanation subjectivity by the use of mathematics method such as information entropy? Key words: oil gas concentration area; sedimentary micro facies； independent component analysis（ICA）; seismic attribute; aliasing signal; information entropy 基金项目：国家科技攻关项目(2004BA616A07)资助 通信作者：王任一，男，1968年9月生，1991年毕业于成都地质学院，现为中国地质大学(武 汉)在读博士研究生，主要从事油气［□开发研究工作。E-mail:wry2899@ 163.com 独立分量分析（Independent Component Analysis, ICA）是上世纪九十年代末新发展 起来的一种全新数据分析处理技术