Open AccessStudy on Real-time Lithology Identification Method of Logging-while-drilling
Author(s) -
Gang Chen
Publication year - 2020
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/546/5/052007
Subject(s) - lithology , identification (biology) , support vector machine , geology , data mining , computer science , sample (material) , interpretation (philosophy) , well logging , pattern recognition (psychology) , artificial intelligence , petroleum engineering , petrology , chemistry , botany , chromatography , biology , programming language
In geophysical exploration and development, lithology identification is the basis for understanding the strata and solving the reservoir parameters. Through the comparative study of three popular lithology identification methods including SVM, GRNN and Elman, each method in the experiments takes the same logging data as a sample to predict and identify, and the most suitable identification method is selected and applied to the automatic identification of lithologic interpretation. It is found that the prediction accuracy of SVM is highest in automatic lithology identification, and the method is suitable for computer automatic analysis in terrestrial interpretation system.