基于非线性变时窗相干算法的不连续性检测方法  

Subtle discontinuity detection with nonlinear variable-time window coherency algorithm

作　　者：蔡涵鹏[1] 胡光岷[1] 贺振华[2] 王梦 贺锡雷[2] 

机构地区：[1]电子科技大学资源与环境学院,四川成都611731 [2]成都理工大学复杂油气藏地球物理研究所,四川成都610059 [3]中国石油川庆钻探工程有限公司地球物理勘探公司,四川成都610213

出　　处：《石油地球物理勘探》2016年第2期371-375,209,共5页Oil Geophysical Prospecting

基　　金：电子科技大学科研启动基金项目(ZYGX2015KYQD049);国家自然科学基金项目(41374111;41374134)联合资助

摘　　要：为了提高相干体技术刻画地震数据体中地层细微不连续性特征的精度,提出了一种基于非线性变时窗的相干算法。与基于本征结构的相干体技术C3相比,该方法包括三项关键技术:1利用时频分析方法计算的瞬时中心频率自适应地确定相干体属性的时窗长度;2应用Kendall秩相关分析代替Pearson相关分析构建用于计算相干属性的正定矩阵;3采用信息散度作为分析点相干度量。实际资料应用结果表明:此法计算的相干体属性不仅压制了沿同相轴分布的低相干条带假同相轴,取得了较好的地层细微不连续性检测效果,而且计算效率得到显著的提高。To enhance coherence cube in subtle discontinuity characterization on seismic data,we propose a nonlinear variable-time window coherence algorithm. Compared with the conventional eigen structure-based coherence computations(C3),the proposed algorithm includes three key parts:A.The instantaneous center frequency,calculated via time-frequency analysis,is used to determine adaptively time-window length for calculating coherence attribute;B.Kendall rank correlation analysis,instead of Pearson correlation analysis,is employed to construct the positive definite matrix for calculating coherence attribute;C.Information divergence is defined as coherence measurement of the analysis point.Experiment results show that the proposed algorithm suppress low-coherence strip phenomenon along seismic event,enhance 3D structural and stratigraphic subtle discontinuity characterization,and improve calculation efficiency.

关 键 词：相干体 不连续性 Kendall秩相关分析 瞬时中心频率 信息散度 

分 类 号：P631.44[天文地球—地质矿产勘探]