摘要
采用主成分分析与判别分析相结合的方法对测井曲线进行自动分层。首先对所有井进行数据预处理,通过程序比较字符串的方式提取出其中的公共曲线,然后对处理过的数据用R编程实现主成分分析,选择主成分代替原来的数据,达到降维的目的。最后以标准井提供的分层结果作为样本进行距离判别分析,对剩下井进行自动分层,并与手工分层的结果比对,给出最终处理的分层结果。
The principal component analysis and discriminated analysis are applied to automatically stratify well logging curves. Firstly, the data preprocessing to all wells is done. By the way of comparing the character string, its common curves are picked up. Then, principal component analysis to the processed data is realized by using R pro- gramming. Principal component is used to replace primary data so as to reduce dimension. Finally, discriminated analysis is done based on the sample of stratified result supplied by standard well. The rest wells are automatically stratified. By comparing with the result of manual stratification, the final - processed stratified result is obtained.
出处
《黑龙江大学自然科学学报》
CAS
北大核心
2012年第3期322-326,共5页
Journal of Natural Science of Heilongjiang University
关键词
测井曲线
主成分分析
马氏距离
判别分析
well logging curves
principal component analysis
Mahalanobis distances
discriminated analysis
