大小模型协同的油气田开发智能体架构设计与展望

Architecture design and outlook of intelligent agents for oil and gas field development based on large-small model collaboration

摘要

油气田开发智能化面临多源异构数据处理、领域知识融合、复杂任务推理与流程自动化等需求,大小模型协同为油气智能体的工程化构建提供重要支撑。系统梳理大模型技术演进脉络、工业大模型核心能力及其在油气行业的应用进展,分析油气开发场景的关键能力需求,提出面向油气任务闭环的工程化技术路径。围绕自动历史拟合、油藏生产优化、井筒举升调控、视频监控与集输调控等典型任务,归纳油气智能体在任务组织、工具链协同与模块化实现方面的主要路径,形成适用于多任务场景的架构方案与协同机制。结合X油田O1-B2区块自动历史拟合与油藏生产优化案例,说明相关闭环组织流程及工程应用模式。最后,从多模态理解、高可信解释机制、任务迁移泛化能力与知识演化闭环等方面展望未来研究方向,为油气智能系统的持续演进提供理论支持与工程参考。

Abstract

Intelligent oil and gas field development is facing increasing demands for multi-source heterogeneous data processing, domain knowledge fusion, complex task reasoning and workflow automation. Large-small model collaboration can provide important supports for the engineering construction of oil and gas intelligent agents. In this paper, the technological evolution and main capabilities of industrial large models and their application progress in the oil and gas industry were systematically reviewed, and the key capability requirements of oil and gas development scenarios were analyzed. An engineering technical path oriented to the closed loop of oil and gas tasks was proposed. Focusing on typical tasks, such as automatic history matching, reservoir production optimization, wellbore artificial lift control, video monitoring and gathering and transportation regulations, the main paths of oil and gas intelligent agents in task organization, toolchain collaboration and modular implementation were summarized, and an architecture scheme and collaboration mechanism applicable to multi-task scenarios were formed. Combined with a case study of automatic history matching and reservoir production optimization in block O1-B2 of X Oilfield, the related closed-loop organization process and engineering application mode were illustrated. Finally, future research directions were discussed from the perspectives of multimodal understanding, highly trustworthy interpretation mechanisms, task transfer and generalization capability and closed-loop knowledge evolution, providing theoretical support and engineering reference for the continuous evolution of oil and gas intelligent systems.

关键词

大语言模型 / 油气田开发 / 智能体 / 油藏动态分析 / 智能决策

Key words

large language model / oil and gas field development / intelligent agent / reservoir dynamic analysis / intelligent decision-making

参考文献

[1] 李欣, 王洪亮, 闫林, 等 . 油气人工智能技术发展综述与体系框架构想[J]. 国际石油经济, 2025, 33(5): 10-18.

[2] LI Xin, WANG Hongliang, YAN Lin, et al. Review of AI technology development in oil and gas fields and framework architecture conceptualization[J]. International Petroleum Economics, 2025, 33(5): 10-18.

[3] GONG R, DING Y, WANG Z, et al. A survey of low-bit large language models: basics, systems, and algorithms[J]. Neural Networks, 2025, 192: 107856.

[4] CAFFAGNI D, COCCHI F, BARSELLOTTI L, et al. The revolution of multimodal large language models: a survey[C]// Findings of the Association for Computational Linguistics: ACL 2024, Bangkok, Thailand 2024: 13590-13618.

[5] YIN S, FU C, ZHAO S, et al. A survey on multimodal large language models[J]. National Science Review, 2024, 11(12): nwae403.

[6] 宫敬, 吴冕, 赵周丙, 等 . 油气管网行业大模型的思考及应用探索[J]. 油气储运, 2025, 44(4): 379-393.

[7] GONG Jing, WU Mian, ZHAO Zhoubing, et al. Reflections and application exploration of large models for the oil and gas pipeline network industry[J]. Oil & Gas Storage and Transportation, 2025, 44(4): 379-393.

[8] 潘焕泉, 刘剑桥, 龚斌, 等 . 油藏动态分析场景大模型构建与初步应用[J]. 石油勘探与开发, 2024, 51(5): 1175-1182.

[9] PAN Huanquan, LIU Jianqiao, GONG Bin, et al. Construction and preliminary application of large language model for reservoir performance analysis[J]. Petroleum Exploration and Development, 2024, 51(5): 1175-1182.

[10] 吴耀辉, 刘梅全, 李雪松, 等 . 油气大模型在钻完井领域的应用研究现状及展望[J]. 钻井液与完井液, 2025, 42(2): 143-154.

[11] WU Yaohui, LIU Meiquan, LI Xuesong, et al. The status-quo of the application and research and prospect of oil and gas large model in well drilling and completion industry[J]. Drilling Fluid & Completion Fluid, 2025, 42(2): 143-154.

[12] 杨明澔, 李小波, 刘兴邦, 等 . 人工智能大模型在油气勘探开发领域的应用及挑战[J]. 石油科技论坛, 2024, 43(6): 107-113, 125.

[13] YANG Minghao, LI Xiaobo, LIU Xingbang, et al. Application of artificial intelligence pretrained foundation models in exploration and development and challenges in this area[J]. Oil Forum, 2024, 43(6): 107-113,125.

[14] 刘合, 任义丽, 李欣, 等 . 油气行业人工智能大模型应用研究现状及展望[J]. 石油勘探与开发, 2024, 51(4): 910-923.

[15] LIU He, REN Yili, LI Xin, et al. Research status and application of artificial intelligence large models in the oil and gas industry[J]. Petroleum Exploration and Development, 2024, 51(4): 910-923.

[16] 刘合, 屈如意, 王素玲, 等 . 人工智能驱动多源融合在油气行业的探索与应用[J]. 中国石油大学学报(自然科学版), 2025, 49(4): 124-134.

[17] LIU He, QU Ruyi, WANG Suling, et al. Artificial intelligence-driven multi-source integration: exploration and application in oil and gas industry[J]. Journal of China University of Petroleum (Edition of Natural Science), 2025, 49(4): 124-134.

[18] GOU J, YU B, MAYBANK S J, et al. Knowledge distillation: a survey[J]. International Journal of Computer Vision, 2021, 129(6): 1789-1819.

[19] GAO C, LAN X, LI N, et al. Large language models empowered agent-based modeling and simulation: a survey and perspectives[J]. Humanities and Social Sciences Communications, 2024, 11: 1259.

[20] BORGHI A M, DE LIVIO C, GERVASI A M, et al. Language as a cognitive and social tool at the time of large language models[J]. Journal of Cultural Cognitive Science, 2024, 8(3): 179-198.

[21] AFTAN S, SHAH H . A survey on BERT and its applications[C]// 2023 20th Learning and Technology Conference (L&T), Jeddah, Saudi Arabia, 2023: 161-166.

[22] HANNA M, LIU O, VARIENGIEN A . How does GPT-2 compute greater-than?: interpreting mathematical abilities in a pre-trained language model[C]// Advances in Neural Information Processing Systems 36, New Orleans, Louisiana, USA, 2023: 76033-76060.

[23] GARDAZI N M, DAUD A, MALIK M K, et al. BERT applications in natural language processing: a review[J]. Artificial Intelligence Review, 2025, 58(6): 166.

[24] CHEN H, XIANG Q, HU J, et al. Comprehensive exploration of diffusion models in image generation: a survey[J]. Artificial Intelligence Review, 2025, 58(4): 99.

[25] WANG X, CHEN G, QIAN G, et al. Large-scale multi-modal pre-trained models: a comprehensive survey[J]. Machine Intelligence Research, 2023, 20(4): 447-482.

[26] MAHOWALD K, IVANOVA A A, BLANK I A, et al. Dissociating language and thought in large language models[J]. Trends in Cognitive Sciences, 2024, 28(6): 517-540.

[27] LIU P, YUAN W, FU J, et al. Pre-train, prompt, and predict: a systematic survey of prompting methods in natural language processing[J]. ACM Computing Surveys, 2023, 55(9): 1-35.

[28] YANG J, JIN H, TANG R, et al. Harnessing the power of LLMs in practice: a survey on ChatGPT and beyond[J]. ACM Transactions on Knowledge Discovery from Data, 2024, 18(6): 1-32.

[29] CHANG Y, WANG X, WANG J, et al. A survey on evaluation of large language models[J]. ACM Transactions on Intelligent Systems and Technology, 2024, 15(3): 1-45.

[30] XI Z, CHEN W, GUO X, et al. The rise and potential of large language model based agents: a survey[J]. Science China Information Sciences, 2025, 68(2): 121101.

[31] LU W, LUU R K, BUEHLER M J . Fine-tuning large language models for domain adaptation: exploration of training strategies, scaling, model merging and synergistic capabilities[J]. npj Computational Materials, 2025, 11: 84.

[32] OVADIA O, BRIEF M, MISHAELI M, et al. Fine-tuning or retrieval? comparing knowledge injection in LLMs[C]// Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, Miami, Florida, USA, 2024: 237-250.

[33] LI Z, CHEN C, LI M, et al. Exploring formal defeasible reasoning of large language models: a chain-of-thought approach[J]. Knowledge-Based Systems, 2025, 319: 113564.

[34] 匡立春, 刘合, 任义丽, 等 . 人工智能在石油勘探开发领域的应用现状与发展趋势[J]. 石油勘探与开发, 2021, 48(1): 1-11.

[35] KUANG Lichun, LIU He, REN Yili, et al. Application and development trend of artificial intelligence in petroleum exploration and development[J]. Petroleum Exploration and Development, 2021, 48(1): 1-11.

[36] 贾承造. 中国石油工业上游发展面临的挑战与未来科技攻关方向[J]. 石油学报, 2020, 41(12): 1445-1464.

[37] JIA Chengzao. Development challenges and future scientific and technological researches in China̍s petroleum industry upstream[J]. Acta Petrolei Sinica, 2020, 41(12): 1445-1464.

[38] 李阳, 廉培庆, 薛兆杰, 等 . 大数据及人工智能在油气田开发中的应用现状及展望[J]. 中国石油大学学报(自然科学版), 2020, 44(4): 1-11.

[39] LI Yang, LIAN Peiqing, XUE Zhaojie, et al. Application status and prospect of big data and artificial intelligence in oil and gas field development[J]. Journal of China University of Petroleum (Edition of Natural Science), 2020, 44(4): 1-11.

[40] MUSTAPHA K B . A survey of emerging applications of large language models for problems in mechanics, product design, and manufacturing[J]. Advanced Engineering Informatics, 2025, 64: 103066.

[41] NIE T, SUN J, MA W . Exploring the roles of large language models in reshaping transportation systems: a survey, framework, and roadmap[J]. Artificial Intelligence for Transportation, 2025, 1: 100003.

[42] ZHANG L, CHEN Z . Opportunities of applying large language models in building energy sector[J]. Renewable and Sustainable Energy Reviews, 2025, 214: 115558.

[43] ZHOU X G, JING X W, WU W K, et al. Preliminary research on applications of large language models in E&P industry[R]. SPE 220833-MS, 2024.

[44] MA Z, SANTOS J E, LACKEY G, et al. Information extraction from historical well records using a large language model[J]. Scientific Reports, 2024, 14: 31702.

[45] OGUNDARE O, MADASU S, WIGGINS N . Industrial engineering with large language models: a case study of ChatGPT̍s performance on oil & gas problems[C]// 2023 11th International Conference on Control, Mechatronics and Automation (ICCMA), Grimstad, Norway, 2024: 458-461.

[46] ALTYNOVA A Y, KOZHEVIN A A, DUBOVIK A S, et al. Advancing geoscience with multi-modal AI: a comprehensive copilot[R]. SPE 222053-MS, 2024.

[47] JACINTO M, SILVA M, OLIVEIRA L, et al. Geologist copilot: generative AI application to support drilling operations with automated and explained lithology interpretation[C]// 85th EAGE Annual Conference & Exhibition, Oslo, Norway, 2024.

[48] CHEN Z, ZHANG Y, LI J, et al. Artificial intelligence large model for logging curve reconstruction[J]. Petroleum Exploration and Development, 2025, 52(3): 842-854.

[49] SHEKHAWAT D, BARUA J, BHATIA K, et al. Offshore production surveillance and intervention using multi agent AI[R]. SPE 226728-MS, 2025.

[50] WIEGAND K, BEDEWI M, MUKUNDAKRISHNAN K, et al. Using generative AI to build a reservoir simulation assistant[R]. SPE 221987-MS, 2024.

[51] ZHAO X, WEI S, WANG L, et al. Large language model empowered automated reservoir agent: a win-win strategy for reservoir intelligent management and transformation[R]. SPE 229646-MS, 2025.

[52] AL-RBEAWI S . A review of modern approaches of digitalization in oil and gas industry[J]. Upstream Oil and Gas Technology, 2023, 11: 100098.

[53] 张菲菲, 王茜, 王学迎, 等 . 油气井工程多源多模态数据融合技术与展望[J]. 天然气工业, 2024, 44(9): 152-166.

[54] ZHANG Feifei, WANG Qian, WANG Xueying, et al. Multi-source and multi-modal data fusion technology and its prospect in oil and gas well engineering[J]. Natural Gas Industry, 2024, 44(9): 152-166.

[55] LIU Z, GUDALA M, YAN B . Efficient optimization of coupled geothermal reservoir modeling and power plant off-design based on deep learning[J]. Advances in Geo-Energy Research, 2025, 18(1): 84-98.

[56] 辛国靖, 张凯, 田丰, 等 . 基于概率建模的分层产液劈分方法[J]. 中国石油大学学报(自然科学版), 2024, 48(2): 109-117.

[57] XIN Guojing, ZHANG Kai, TIAN Feng, et al. Prediction method of fluid production profiles based on a probabilistic modeling method[J]. Journal of China University of Petroleum (Edition of Natural Science), 2024, 48(2): 109-117.

[58] ZHU Y . A knowledge graph and BiLSTM-CRF-enabled intelligent adaptive learning model and its potential application[J]. Alexandria Engineering Journal, 2024, 91: 305-320.

[59] BORHANIFARD Z, FAILI H, YAGHOOBZADEH Y . Combining replay and LoRA for continual learning in natural language understanding[J]. Computer Speech & Language, 2025, 90: 101737.

[60] QU C, DAI S, WEI X, et al. Tool learning with large language models: a survey[J]. Frontiers of Computer Science, 2025, 19(8): 198343.

[61] GUO T, CHEN X, WANG Y, et al. Large language model based multi-agents: a survey of progress and challenges[C]// Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence, 2024: 8048-8057.

[62] ZHANG J, ZHANG K, LIU P, et al. Deep Bayesian surrogate models with adaptive online sampling for ensemble-based data assimilation[J]. Journal of Hydrology, 2025, 649: 132457.

[63] ZHANG J, KANG J, ZHANG K, et al. An efficient transformer-based surrogate model with end-to-end training strategies for automatic history matching[J]. Geoenergy Science and Engineering, 2024, 240: 212994.

[64] DAI Q, ZHANG L, ZHANG K, et al. Carbon-pricing optimization framework for oilfield carbon dioxide-water alternation with economic and environmental gains[J]. Energy Conversion and Management, 2026, 348: 120613.

[65] DAI Q Y, ZHANG L M, ZHANG K, et al. Integrated optimization of reservoir production and layer configurations using relational and regression machine learning models[J]. Petroleum Science, 2025, 22(9): 3745-3759.

[66] ZHOU X, ZHANG L, QI J, et al. Integrated wellbore-surface pressure control production optimization for shale gas wells[J]. Natural Gas Industry B, 2025, 12(2): 123-134.

[67] WANG X, ZHANG L, WANG Y, et al. Visual-Language contrastive learning for zero-shot compound fault diagnosis in sucker rod wells[J]. Measurement, 2025, 243: 116320.

[68] RAMACHANDRA B, JONES M J, VATSAVAI R R . A survey of single-scene video anomaly detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 44(5): 2293-2312.

[69] MONTAGNA A F, CAFARO D C, GROSSMANN I E, et al. Surface facility optimization for combined shale oil and gas development strategies[J]. Optimization and Engineering, 2023, 24(4): 2321-2355.

基金资助

国家自然科学基金项目(52325402)

国家自然科学基金项目(52441411)

国家自然科学基金项目(52274057)

地球深部探测与矿产资源勘查国家科技重大专项(2024ZD1004302-04)

国家重点研发计划(2023YFB4104200)