重庆科技学院学报（自然科学版）2024，Vol.26Issue(3)：15-20,6.DOI:10.19406/j.issn.1673-1980.2024.03.003
基于测录井资料多元统计分析识别浅层气
Identification of Shallow Gas Based on Multivariate Statistical Analysis of Logging Data
摘要
Abstract
A quantitative identification method for shallow gas is explored by comprehensively utilizing logging statis-tical data.A classification discrimination model based on Fisher discrimination and binary Logistic regression is es-tablished to distinguish the fluid properties of unknown samples.Multiple logging curves closely related to reservoir gas content are selected and shallow gas advantage discrimination curves are reconstructed.Based on the above two discrimination classification curves,a shallow gas discrimination chart is established and a shallow gas discrimina-tion method is formed.This method is used for drilling shallow gas in Bohai P oilfield,and the success rate of shal-low gas discrimination is greater than 80%.Its effectiveness and pertinence have been verified.关键词
浅层气/测井资料/气测录井/数理统计/Fisher判别法/Logistics回归法Key words
shallow gas/well logging data/gas logging/mathematical statistics/Fisher discrimination/Logistic regression分类
能源科技引用本文复制引用
关叶钦,许赛男,时新磊,陆云龙..基于测录井资料多元统计分析识别浅层气[J].重庆科技学院学报（自然科学版）,2024,26(3):15-20,6.基金项目
中海石油(中国)有限公司科技项目"渤海油田3000万t持续稳产关键技术研究"(CNOOC-KJ135-ZDXM36-TJ-03) （中国）
