摘要
根据测井曲线数据,运用移动平均方法实现自动分层,再对得到的分层数据,采用交会图、BP神经网络、支持向量机这三种岩性识别方法,对比分析每种方法使用相同的测井资料作为样本,对相同的测井数据做预测分析,与此同时与MATLAB识别结果进行对比,从中选出比较合适的岩性识别方法。
According to the logging curve data,the moving average method is used to realize the automaticstratification,and then the hierarchical data are obtained.The three lithology identification methods,such as the intersection map,BP neural network and support vector machine,are used to compare and analyze each method using the same Logging data as a sample,the same logging data to do the forecast analysis,at the same time with the MATLAB recognition results were compared,from the selection of more appropriate lithology identification method.
作者
吴非
王卫
王佳琦
Wu Fei;Wang Wei;Wang Jiaqi(Sinopec Research Institute of Petroleum Engineering,Beijing 100101;Beijing Science and technology innovation research center,Beijing 100744)
出处
《科技风》
2021年第15期18-20,共3页
基金
国家科技重大专项课题“低渗透储层高精度随钻成像技术研究”(2016ZX05021-002)。
关键词
测井数据
曲线分层
交会图
BP神经网络
支持向量机
岩性识别
logging curves
curve layering
cross plot
BP Neural Network
support vector machine
lithology identification
