A novel approach to classify lithology of reservoir formations using GrowNet and Deep-Insight with physic-based feature augmentation

IF 3.5 3区 工程技术 Q3 ENERGY & FUELS Energy Science & Engineering Pub Date : 2024-10-23 DOI:10.1002/ese3.1895
Seyed Hamid Reza Mousavi, Seyed Mojtaba Hosseini-Nasab
{"title":"A novel approach to classify lithology of reservoir formations using GrowNet and Deep-Insight with physic-based feature augmentation","authors":"Seyed Hamid Reza Mousavi,&nbsp;Seyed Mojtaba Hosseini-Nasab","doi":"10.1002/ese3.1895","DOIUrl":null,"url":null,"abstract":"<p>Manual interpretation of geophysical logging data can be a tedious and time-consuming task in the case of the nonlinear behavior of well-logging signals. This study aims to enhance lithology classification of reservoir formations through advanced machine learning (ML) and deep learning (DL) techniques, introducing and comparing three novel algorithms, GrowNet, Deep-Insight, and blender, against traditional models like random forest (RF) and support vector machine (SVM). Data from the South and North Viking Graben regions, encompassing 12 lithological facies, was preprocessed through cleaning, normalization, transformation, and imputation of missing values using regression models. The data set was enhanced with physic-based features and balanced using SMOTE and NearMiss algorithms. Deep-Insight converted tabular data into images for a convolutional neural network (CNN), significantly improving classification accuracy compared to conventional models such as decision trees (DTs). GrowNet and blender models leveraged hybrid approaches for enhanced performance. These hybrid approaches successfully addressed data imbalance and enhanced model learning, outperforming classic methods. The GrowNet and blender models for lithology classification successfully increased the penalty score and accuracy compared to the FORCE 2020 competition. Additionally, introducing the class prediction error plot visualizes multiclass classification performance more effectively than using a confusion matrix. These novel models in multiclass classification contribute to the petroleum industry by providing more accurate and reliable lithology classification, thereby improving reservoir characterization and exploration efficiency.</p>","PeriodicalId":11673,"journal":{"name":"Energy Science & Engineering","volume":null,"pages":null},"PeriodicalIF":3.5000,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/ese3.1895","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Energy Science & Engineering","FirstCategoryId":"5","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/ese3.1895","RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ENERGY & FUELS","Score":null,"Total":0}
引用次数: 0

Abstract

Manual interpretation of geophysical logging data can be a tedious and time-consuming task in the case of the nonlinear behavior of well-logging signals. This study aims to enhance lithology classification of reservoir formations through advanced machine learning (ML) and deep learning (DL) techniques, introducing and comparing three novel algorithms, GrowNet, Deep-Insight, and blender, against traditional models like random forest (RF) and support vector machine (SVM). Data from the South and North Viking Graben regions, encompassing 12 lithological facies, was preprocessed through cleaning, normalization, transformation, and imputation of missing values using regression models. The data set was enhanced with physic-based features and balanced using SMOTE and NearMiss algorithms. Deep-Insight converted tabular data into images for a convolutional neural network (CNN), significantly improving classification accuracy compared to conventional models such as decision trees (DTs). GrowNet and blender models leveraged hybrid approaches for enhanced performance. These hybrid approaches successfully addressed data imbalance and enhanced model learning, outperforming classic methods. The GrowNet and blender models for lithology classification successfully increased the penalty score and accuracy compared to the FORCE 2020 competition. Additionally, introducing the class prediction error plot visualizes multiclass classification performance more effectively than using a confusion matrix. These novel models in multiclass classification contribute to the petroleum industry by providing more accurate and reliable lithology classification, thereby improving reservoir characterization and exploration efficiency.

Abstract Image

查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
利用基于物理特征增强的 GrowNet 和 Deep-Insight 对储层岩性进行分类的新方法
在测井信号具有非线性行为的情况下,人工解释地球物理测井数据可能是一项乏味且耗时的任务。本研究旨在通过先进的机器学习(ML)和深度学习(DL)技术加强储层岩性分类,引入了三种新型算法:GrowNet、Deep-Insight 和 blender,并与随机森林(RF)和支持向量机(SVM)等传统模型进行了比较。来自南维京海盆和北维京海盆地区的数据包含 12 个岩性面,通过清理、归一化、转换和使用回归模型估算缺失值进行了预处理。数据集使用物理特征进行了增强,并使用 SMOTE 和 NearMiss 算法进行了平衡。Deep-Insight 将表格数据转换为卷积神经网络 (CNN) 的图像,与决策树 (DT) 等传统模型相比,显著提高了分类准确性。GrowNet 和 blender 模型利用混合方法提高了性能。这些混合方法成功地解决了数据不平衡问题,增强了模型学习能力,表现优于传统方法。与 FORCE 2020 竞赛相比,用于岩性分类的 GrowNet 和 blender 模型成功提高了罚分和准确率。此外,与使用混淆矩阵相比,引入类预测误差图能更有效地可视化多类分类性能。这些新颖的多类分类模型为石油工业做出了贡献,提供了更准确可靠的岩性分类,从而提高了储层特征描述和勘探效率。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
Energy Science & Engineering
Energy Science & Engineering Engineering-Safety, Risk, Reliability and Quality
CiteScore
6.80
自引率
7.90%
发文量
298
审稿时长
11 weeks
期刊介绍: Energy Science & Engineering is a peer reviewed, open access journal dedicated to fundamental and applied research on energy and supply and use. Published as a co-operative venture of Wiley and SCI (Society of Chemical Industry), the journal offers authors a fast route to publication and the ability to share their research with the widest possible audience of scientists, professionals and other interested people across the globe. Securing an affordable and low carbon energy supply is a critical challenge of the 21st century and the solutions will require collaboration between scientists and engineers worldwide. This new journal aims to facilitate collaboration and spark innovation in energy research and development. Due to the importance of this topic to society and economic development the journal will give priority to quality research papers that are accessible to a broad readership and discuss sustainable, state-of-the art approaches to shaping the future of energy. This multidisciplinary journal will appeal to all researchers and professionals working in any area of energy in academia, industry or government, including scientists, engineers, consultants, policy-makers, government officials, economists and corporate organisations.
期刊最新文献
Issue Information Similar simulation test of the mechanical properties of layered composite rock mass A novel approach to classify lithology of reservoir formations using GrowNet and Deep-Insight with physic-based feature augmentation Combined genetic algorithm and response surface methodology-based bi-optimization of a vertical-axis wind turbine numerically simulated using CFD Experimental study on the utilization of Fly ash solid waste in tunnel shotcrete materials
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1