论文部分内容阅读
油水层判断是测井地质解释工作中最主要的工作之一。为了提高测井解释质量,现已用计算机进行多元判别分析。二元判别分析是基本的判别分析。本文提出用归属组的概率选择关键资料点及剔掉某些测井参数范围内过于分散的资料点的办法,使得二元判别分析确定的分界限效果与人工确定的相同,并保证多元判别确定的分界限比人工确定的更精确。
Judgment of oil-water layer is one of the most important work in logging geologic interpretation. In order to improve the quality of logging interpretation, a computer has been used to conduct multiple discriminant analysis. Binary discriminant analysis is the basic discriminant analysis. This paper proposes to select the key data points by the probability of belonging group and to delete the data points scattered too far in the range of some well logging parameters so that the bounded discriminant effect of bivariate discriminant analysis is the same as that determined by hand and the multivariate discriminant determination The delimitation is more accurate than the manual determination.