摘要
通过对CCSD主孔100~2000m井段变质岩测井资料以及岩芯资料的综合分析,结合多元统计学方法,建立了一套用Fisher判别变质岩的二级模型。利用100~1000m井段的岩性测井资料建立模型,对1000~2000m井段的岩性进行识别,一级大类变质岩的识别准确率为89.68%,二级片麻岩的识别准确率为92.96%,二级榴辉岩的识别准确率为84.38%,取得了良好的应用效果,与单一的识别方法相比,提高了准确率。
Based on the logging and coring data from depths 100 to 2000 meters in CCSD Main Hole,as well as multivariate statistical methods,we establish a metamorphic rocks recognition model by Fisher discriminant which has two levels.We build the models to identify the lithology from 1000 to 2000 meters with the logging data from 100 to 1000 meters: the first level of metamorphic rock identification rate is 89.68%,the second level of gneiss identification rate is 92.26%,and the second level of eclogite identification rate is 84.38%.Compared to a single identification method,the effect of its application is achieved and the accuracy is improved.