文献类型：会议论文

英文题名：Research on automatic evaluation model of carbonate reservoirs in Beibu Gulf

作者：Zhang, Ying[1]; Li, Zixin[1]; Peng, Xiaohong[1]; Li, Zhentao[1]

机构：[1] School of Mathematics and Computer Science, Guangdong Ocean University, Guangdong, Zhanjiang, China

会议论文集：2023 3rd Guangdong-Hong Kong-Macao Greater Bay Area Artificial Intelligence and Big Data Forum, AIBDF 2023

会议日期：September 22, 2023 - September 24, 2023

会议地点：Guangzhou, China

语种：英文

外文关键词：Adaptive boosting - Carbonation - Forestry - Learning systems - Petroleum reservoir evaluation

外文摘要：In the context of the Weixinan Depression’s carbonate reservoirs, this study sought to address the scientific challenges of automating reservoir evaluations. Leveraging machine learning theory, and employing a curated dataset from the depression, we embarked on constructing a model tailored for carbonate reservoir evaluations. Three machine learning algorithms apt for handling multi-feature, multi-classification complexities Support Vector Machines, Gradient Boosting Trees, and Random Forests were meticulously selected. A pivotal advancement in this research was the use of a neural network approach to effectively interpolate missing data in the drilling logs, ensuring enhanced data integrity. After rigorous evaluations, the Gradient Boosting Tree-based model emerged as the most proficient, yielding an accuracy of 86.26%, positioning it as an effective methodology for comprehensive reservoir classification utilizing rock physics indicators. ? 2023 Copyright held by the owner/author(s).