基于面波频散曲线聚类分析的近地表横波速度反演  

Near surface S-wave velocity inversion based on cluster analysis of surface wave dispersion curve

作　　者：宋欣悦 刘国昌[1] 杜婧 迟麟 王志勇 时岚婷 王依萌 王兴宇 SONG XinYue;LIU GuoChang;DU Jing;CHI Lin;WANG ZhiYong;SHI LanTing;WANG YiMeng;WANG XingYu(State Key Laboratory of Petroleum Resources and Prospecting,China University of Petroleum(Beijing),Beijing 102249,China;ZKCC Marine Information System Co.,Ltd.,Beijing 100190,China;Institute of Geophysical and Geochemical Exploration,Chinese Academy of Geological Sciences,Langfang 065000,China;Exploration and Development Research Institute of Daqing Oilfield Company Ltd.,Daqing 163712,China)

机构地区：[1]中国石油大学(北京)油气资源与探测国家重点实验室,北京102249 [2]中科长城海洋信息系统有限公司,北京100190 [3]中国地质科学院地球物理地球化学勘查研究所,廊坊065000 [4]大庆油田有限责任公司勘探开发研究院,大庆163712

出　　处：《地球物理学报》2023年第7期3026-3047,共22页Chinese Journal of Geophysics

基　　金：国家自然科学基金项目(42074128);中国石油天然气集团有限公司科技管理部项目"物探应用基础实验和前沿理论方法研究"(2022DQ0604-02);中国石油大学(北京)科研基金(2462020YXZZ006);中国石油天然气集团有限公司-中国石油大学(北京)战略合作科技专项(ZLZX2020-03)联合资助。

摘　　要：获取准确的近地表横波速度对复杂地表条件下弹性波地震数据处理和成像非常重要.在浅层面波工程勘探中通过反演提取的频散曲线可以获得近地表横波速度结构.在多道面波频散曲线分析中,频散关系拾取的精度直接影响速度反演结果的可靠性.本文在多道面波叠加及自动拾取频散曲线基础上,提出了基于面波频散曲线聚类分析的近地表横波速度反演方法.该方法充分考虑了低信噪比条件下面波频散曲线的不确定性,通过在频散曲线拾取中引入曼哈顿距离K-Means聚类算法提高频散曲线拾取的准确性.采用多道多窗口叠加技术提高了面波反演对横向速度变化的适应性,通过聚类算法和多窗口叠加提高反演的可靠性,聚类算法获得较准确的频散曲线更利于后续横波速度反演过程.模拟数据算例对比表明本文提出的方法比常规算法效果更好,精度更高.将提出的方法应用于工程勘探和油气勘探的面波数据反演中,结果也验证了该方法的有效性.Obtaining accurate near surface shear wave velocity is very important for elastic wave seismic data processing and imaging under complex surface conditions.In shallow layer wave engineering exploration,the near surface shear wave velocity structure can be obtained by inversion of the dispersion curve.In the analysis of multi-channel surface wave dispersion curve,the accuracy of frequency dispersion relationship directly affects the reliability of velocity inversion results.Based on multi-channel surface wave superposition and automatic pick-up of dispersion curves,a near surface shear wave velocity inversion method based on cluster analysis of surface wave dispersion curves is proposed in this paper.This method fully considers the uncertainty of wave dispersion curve under the condition of low signal-to-noise ratio,and improves the accuracy of dispersion curve picking by introducing Manhattan distance K-Means clustering algorithm into dispersion curve picking.The multi-channel and multi window stacking technology is used to improve the adaptability of surface wave inversion to the variation of transverse velocity.The reliability of inversion is improved by clustering algorithm and multi window stacking.The clustering algorithm obtains more accurate dispersion curve,which is more conducive to the subsequent shear wave velocity inversion process.The comparison of simulation data shows that the proposed method has better effect and higher accuracy than the conventional algorithm.The proposed method is applied to the inversion of surface wave data in engineering exploration and oil and gas exploration,and the results also verify the effectiveness of the method.

关 键 词：面波反演 横波速度 近地表勘探 频散曲线K-Means聚类 

分 类 号：P631[天文地球—地质矿产勘探]