摘要
通过梳理国内外人工智能技术在地球物理勘探(物探)领域中的发展历程、主要研究进展以及发展方向,总结了智能物探的优势和面临的难题,并提出了解决方案。研究结果表明:(1)物探技术在人工智能发展的第2次浪潮中开始与人工智能技术相结合,得益于物探领域数据量的指数级增长、硬件算力的高速发展以及不断出现的新深度学习框架,智能物探技术从早期的机器学习发展为目前的深度学习,在地震资料处理、解释等方面的应用中取得了大量研究成果。(2)目前智能物探技术被广泛应用于标签集的构建、去噪、断裂检测、层位与层序解释、地震相分类和异常体检测、岩性识别与油气藏开发、地震反演成像等方面,大幅提高了工作效率,降低了工作成本,克服了人工交互操作和人工经验的主观性和不可靠性,助力打破传统物探技术瓶颈。(3)智能物探技术的发展面临着缺少公开的标签数据集、缺少解决地球物理领域问题的智能化框架及尚未形成适用于地球物理领域共享的智能化开发平台等难题,可以从解决数据基础、构建智能平台、开展网络架构基础性研究及与应用场景结合等方面着手解决;此外,智能物探技术的发展方向还包含智能地震成像方法研究,储层成像方法研究,油气大数据挖掘、智能风险评估与智能决策以及超算软件装备研发等方面。
By reviewing the development history,main research progress,and development direction of artifi-cial intelligence technology in the field of geophysical exploration(geophysical exploration)both domestically and internationally,the advantages and challenges of intelligent geophysical exploration were summarized,and solutions were proposed.The results show that:(1)Geophysical technology was integrated with artificial intelli-gence technology in the second wave of artificial intelligence development.Thanks to the exponential growth of data volume in the field of geophysical exploration,the rapid development of hardware computing power,and the emergence of new deep learning frameworks,intelligent geophysical technology has developed from early machine learning to current deep learning,and has achieved a large number of research results in seismic data processing and interpretation.(2)At present,intelligent geophysical technology is widely used in the construc-tion of tag sets,denoising,fault detection,layer and sequence interpretation,seismic facies classification and anomaly detection,lithology identification and reservoir development,and seismic inversion imaging,greatly improving work efficiency,reducing work costs,overcoming the subjectivity and unreliability of manual interac-tion and experience,and helping to break the bottleneck of traditional geophysical technology.(3)The develop-ment of intelligent geophysical technology faces challenges such as a lack of publicly available label datasets,a lack of intelligent frameworks to solve problems in the field of geophysics,and the lack of an intelligent development platform suitable for sharing in the field of geophysics.These challenges can be addressed by addressing data infrastructure,building intelligent platforms,conducting basic research on network architecture,and combining it with application scenarios.In addition,the development direction of intelligent geophysical technology also includes the following aspects:research on intelligent seismic imaging methods,reservoir imaging methods,oil and gas big data mining,intelligent risk assessment and intelligent decision-making,and research and development of supercomputing software equipment.
作者
杨午阳
魏新建
李海山
YANG Wuyang;WEI Xinjian;LI Haishan(PetroChina Research Institute of Petroleum Exploration&Development-Northwest,Lanzhou 730020,China)
出处
《岩性油气藏》
CAS
CSCD
北大核心
2024年第2期170-188,共19页
Lithologic Reservoirs
基金
中国石油天然气集团公司前瞻性基础性项目“物探采集处理解释关键技术研究”(编号:2021DJ37)资助。
关键词
智能物探
大数据
人工智能
机器学习
深度学习
标签数据集
深度学习框架
智能处理与解释
地震资料
intelligent geophysical exploration
big data
artificial intelligence
machine learning
deep learning
labeled datasets
deep learning framework
intelligent processing and interpretation
seismic data