摘要
使用Boosting Tree算法,以录井资料和测井资料为基础,优选出自然伽马、自然电位、冲洗带电阻率、侵入带电阻率、原状地层电阻率、密度、补偿中子、声波时差8个对岩性敏感度较高的测井属性,建立岩性识别模型。使用该方法对玛北油田岩石类型齐全的6号井的目的层岩性进行识别,正确率达到89.1%,优于决策树、支持向量机(SVM)等传统的机器学习方法。使用Boosting Tree算法对岩性进行识别也为测井解释提供了新的思路。
Using the Boosting Tree algorithm, and based on mud logging data and wireline logging data, the logging with high lithology sensitivity, including natural gamma, natural potential, flushed zone resistivity, invaded zone resistivity, undisturbed formation resistivity, density, compensated neutron logging and AC logging, are selected to establish lithology identification model. The developed method is used to identify the lithology of target zone of Well No.6 with complete rock types in Mabei Oilfield, and the correct rate reaches 89.1%, which is better than traditional machine learning methods such as decision tree and support vector machine (SVM). The identification of lithology using the Boosting Tree algorithm provides a new idea for logging interpretation.
作者
江凯
王守东
胡永静
浦世照
段航
王政文
JIANG Kai;WANG Shoudong;HU Yongjing;PU Shizhao;DUAN Hang;WANG Zhengwen(College of Geophysics and Information Engineering,China University of Petroleum(Beijing),Beijing 102249,China;State Key Laboratory of Petroleum Resources and Prospecting,Beijing 102249,China;National Engineering Laboratory for Offshore Oil Exploitation,Beijing 102249,China;Exploration and Development Research Institute,PetroChina Xinjiang Oilfield Company,Karamay,Xinjiang 834000,China)
出处
《测井技术》
CAS
CSCD
2018年第4期395-400,共6页
Well Logging Technology
基金
国家科技重大专项(2016ZX05024-001-004)资助