地质统计学变程优化研究  

Variable Range Optimization in Geostatistics

作　　者：郝红瑞 潘懋[1] 刘振[1] 曹凯[1] HAO Hongrui;PAN Mao;LIU Zhen;CAO Kai(School of Earth and Space Sciences,Peking University,Beijing 100871)

机构地区：[1]北京大学地球与空间科学学院,北京100871

出　　处：《计算机与数字工程》2017年第6期1180-1186,共7页Computer & Digital Engineering

基　　金：国家自然科学基金项目“基于语义的多分辨率储层数据组织与管理”(编号:41472113)资助

摘　　要：三维地质属性建模需要采用各种地质属性的样本数据,经过变异函数分析以及合适的插值算法,对未知区域进行相应插值,以获取研究区内的属性分布。为了构建更为精确的属性模型通常在建模过程中需要一定的人机交互。但在以往属性建模中,变程的有效选取一般根据属性建模的结果予以不断调整来获取,而没有一个参考的标准,这就给属性建模的效率和精度造成很大影响。针对这一问题,尝试进行地质统计学变程优化研究,寻求有效变程同相关影响因素之间的理论模型,从而提高属性建模的效率和精度。It is necessary to adopt sample data of various geological attributes for3D geological attribute modeling,the interpo?lation of unknown regions is performed by variogram analysis and proper interpolation algorithm to obtain the distribution of attri?butes in the study area.In order to build more accurate attribute models,a certain amount of human-computer interaction is usuallyneeded during the modeling process.But in the past property modeling,the effective selection of variable range is usually adjustedaccording to the result of attribute modeling,without a reference standard,and this has great influence on the efficiency and preci?sion of attribute modeling.Based on the problems above,this paper attempts to study the optimization of geostatistical variablerange,and seek the theoretical model between effective variable range and related influencing factors,so as to improve modeling ef?ficiency and accuracy.

关 键 词：属性建模 变异函数 有效变程 属性变化 地质体规模 序贯高斯 

分 类 号：P628[天文地球—地质矿产勘探]