基于多信息融合的层次聚类测井曲线自动分层方法  

Automatic stratification method of hierarchical clustering logging curve based on multi-information fusion

作　　者：张景越 肖小玲[1,2] 王鹏飞[1] 向家富 张翔 ZHANG Jingyue;XIAO Xiaoling;WANG Pengfei;XIANG Jiafu;ZHANG Xiang(School of Computer Science,Yangtze University,Jingzhou 434000,China;MOE Key Laboratory of Exploration Technologies for Oil and Gas Resources,Yangtze University,Wuhan 430100,China)

机构地区：[1]长江大学计算机科学学院,湖北荆州434000 [2]油气资源与勘探技术教育部重点实验室(长江大学),湖北武汉430100

出　　处：《断块油气田》2024年第1期42-49,共8页Fault-Block Oil & Gas Field

基　　金：国家自然科学基金项目“碳酸盐岩不同孔隙结构多尺度三维数字岩心建模方法研究”(41674136)、“复杂地质背景下电成像测井层理面检测与产状快速提取方法研究”(41374148)。

摘　　要：随着人工智能的快速发展,机器学习的应用范围越来越广泛,将机器学习的方法用于测井曲线分层可以提高分层效率和精度。在利用测井资料进行岩性识别、沉积相分析等研究时,先要对测井曲线进行分层。文中提出一种基于多信息融合的层次聚类分层方法,实现了对测井曲线的自动分层。首先,采用滤波的方式滤除曲线上的噪点,对数据进行归一化处理,消除量纲的影响;其次,通过特征优选,选择包含较多地层信息的特征曲线,构造一个滤波器,将其中相似性较高的曲线融合,曲线融合的权值通过遗传算法求得;最后,使用层次聚类方法对多信息融合后的测井数据进行划分,将分层结果与人工分层结果进行对比验证。该方法能够提高分层效率,为地质勘探工作提供可靠的分层依据。With the rapid development of artificial intelligence,the application range of machine learning is becoming more and more extensive.The stratification efficiency and accuracy of logging curve can be improved by using machine learning methods.When using logging data for lithology identification,sedimentary facies analysis and other research,the logging curve should be stratified first.In this paper,a hierarchical clustering and stratification method based on multi-information fusion is proposed to realize the automatic stratification of logging curves.Firstly,the noise on the curve is filtered by filtering,and the data is normalized to eliminate the influence of dimensions.Secondly,feature curves containing more stratigraphic information are selected by feature optimization,the filter is constructed and curves with high similarity are fused,and the weight of the curve fusion is obtained by genetic algorithm.Finally,hierarchical clustering method is used to divide the logging data after multi-information fusion,and its stratification results are compared with the manual stratification results.This method can improve the efficiency of stratification and provide a reliable stratification basis for geological exploration.

关 键 词：多信息融合 层次聚类 测井曲线分层 滤波 遗传算法 

分 类 号：TE319[石油与天然气工程—油气田开发工程]