物探化探计算技术2024，Vol.46Issue(2)：146-153,8.DOI:10.3969/j.issn.1001-1749.2024.02.03
基于BO-XGBoost优化方法的砂岩厚度预测方法研究与应用
Research and application of sandstone thickness prediction method based on BO-XGBoost optimization method
摘要
Abstract
Channel sand body prediction plays a vital role in oil and gas exploration,but in actual exploration and develop-ment,the buried depth of complex channel sand body and the complexity of channel structure will make its well seismic rela-tionship poor,leading to low sandstone prediction accuracy.Given this problem,this paper uses the seismic multi-attribute in-formation,gives full play to the advantages of the dense well pattern,and uses the BO-XGBoost sandstone thickness prediction method to achieve the best prediction effect when the validation set proportion is 25%.The prediction effect is better than the conventional SVM and XGBoost methods.The research method can be used for reference in sandstone thickness prediction of similar complex reservoirs.关键词
贝叶斯优化/XGBoost/砂体/厚度预测Key words
bayesian optimization/XGBoost/sand body/thickness prediction分类
地质学引用本文复制引用
刘烜良,张军华,白青林,王福金,刘中伟,焦红岩..基于BO-XGBoost优化方法的砂岩厚度预测方法研究与应用[J].物探化探计算技术,2024,46(2):146-153,8.基金项目
国家自然科学基金项目(42072169) （42072169）
胜利油田项目(30200003-21-ZC0631) （30200003-21-ZC0631）
