测井技术2018，Vol.42Issue(6)：641-646,6.DOI:10.16489/j.issn.1004-1338.2018.06.006
基于图版法决策树在流体识别中的应用
Application of Cross-plot-based Decision Tree Template Method in Fluid Identification
赵倩 1杨斌 1李星 1王少龙 1魏杰1
作者信息
- 1. 成都理工大学能源学院, 四川 成都 610059
- 折叠
摘要
Abstract
The reservoir characteristics of YD oilfield are the coexistence of oil and gas reservoirs.Due to factors such as fluid properties, reservoir physical properties and high formation water salinity, the log response characteristics of fluids are complex.After several experiments, seven well logging and mud logging parameters were selected, and four fluid identification templates were established.Based on the parameters selected by the template and the log response characteristics of different fluid type, the decision tree analysis method is used to optimize the log response values as the characteristic parameters of the fluid, and the fluid identification decision tree model is also established.The interpretation coincidence rate of the model reached 94.32％.This method effectively solves the problems of the template method, such as uncertainty and multiplicity, and can process large amounts of data quickly and efficiently.Through the effective combination of traditional geophysical statistical template method and data mining decision tree method, the accuracy of fluid identification is improved, which provides a reliable basis for the development of fine reservoirs in the next step of the study area and expands the applications of data mining in geophysics.关键词
测井解释/流体识别/交会图版/决策树模型/数据挖掘Key words
log interpretation/fluid identification/cross-plot/decision tree model/data mining分类
能源科技引用本文复制引用
赵倩,杨斌,李星,王少龙,魏杰..基于图版法决策树在流体识别中的应用[J].测井技术,2018,42(6):641-646,6.
