摘要
气测录井解释方法与模型多种多样,但其真正意义上的解释符合率却较低。该文介绍了一种新型的气测录井解释方法的建立过程,即根据贝叶斯准则,用逐步判别分析法建立了气测录井解释模型,模型建立过程中选用样本数334个,数据源为气测全烃及色谱组分分析数据,根据归属最大概率判别油层、气层、油水同层、水层和干层。解释模型的显著性水平为0.1,总体回判分辨率92％。
There are a variety of gas logging interpretation methods and models, but coincidence rate of interpretation is low in reality. The paper introduces establishing process of a new gas logging interpretation method using step by step discrimination, which is according Bayes criterion to set up gas logging interpretation model with the step by step discrimination analysis method. It selects 334 sample data, all data source are total hydrocarbon and chromatographic component's analytical data. According to incorporation maximum probability, it can discriminate oil-bearing formation, gas-bearing formation, oil-water bed, water layer and dry layer. The distinguished level of interpretation is 0.1, overall regression judgement resolution of it is 90％.
关键词
逐步判别法
气测录井
解释模型
判别函数
概率
分辨率
油田
Gas Logging,Interpretation Model,Step By Stey DisCrimination,Analy-sis,Discriminant Function,Probability,Resolution
