PDF(1196 KB)
![](\"/images/email.png\")
针对石油领域传统气量预测模型在捕捉长时间依赖和刻画气井间非线性、多尺度空间交互方面的不足，提出基于预训练机制的时空图气量预测模型.在预训练阶段，通过Transformer模块构建时间序列掩码自编码器，提取油田生产数据的深层时序特征，生成具备全局上下文感知的段级表征，解决传统模型时序特征提取不充分的问题；在预测阶段，依托时空特征融合机制，将上述时序表征与动态图结构学习模块捕获的气井空间依赖相结合，突破传统方法在长期记忆缺失、预定义图结构不完整场景下的性能瓶颈.实验结果显示，模型预测的平均绝对误差降至0.156，和基准模型相比，误差分别降低6.6%，36.3%，26.9%和67.9%，性能显著提升.该成果为石油勘探中的时空数据建模提供了新的深度学习范式，是时空大模型工程化应用的有益探索.
","endNoteUrl_en":"https://jns.nju.edu.cn/EN/article/getTxtFile.do?fileType=EndNote&id=1759","reference":"| [1] | Box G. Box and Jenkins：Time series analysis，forecasting and control∥Mills T C. A Very British Affair：Six Britons and the Development of Time Series Analysis During the 20th Century.London，United Kindom：Palgrave Macmillan，2013：161-215. |
| [2] | Gon?alves I F A， Silva T M D， Barreto A B，et al. Predicting oil field production using the random forest algorithm∥Anais Estendidos da XXXV Conference on Graphics，Patterns and Images.Porto Alegre，Brazil：Sociedade Brasileira de Computa??o，2022：134-139. |
| [3] | Sadegh S， Shadizadeh S R. Reservoir rock permeability prediction using support vector regression in an Iranian oil field. Journal of Geophysics & Engineering，2012，9(3)：336-344. |
| [4] | 孔令维，许立红，刘树辉. 基于大数据分析的机采井组最优产能模型的确定与应用. 石油石化节能，2023，13(5)：49-53. |
| [5] | Rodriguez A X， Salazar D A. Methodology for the prediction of fluid production in the waterflooding process based on multivariate long?short term memory neural networks. Journal of Petroleum Science and Engineering，2022，208(Part E)：109715. |
| [6] | Avula V R， Kapavarapu B B， K M M，et al. Prediction of oil production using optimized gated recurrent unit (GRU)∥2024 International Conference on Recent Advances in Electrical，Electronics，Ubiquitous Communication，and Computational Intelligence. Chennai,India：IEEE，2024：1-6. |
| [7] | Shao Z Z， Zhang Z， Wang F，et al. Pre?training enhanced spatial?temporal graph neural network for multivariate time series forecasting∥Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. New York，NY，USA：Association for Computing Machinery，2022：1567-1577. |
| [8] | Defferrard M， Bresson X， Vandergheynst P. Convolutional neural networks on graphs with fast localized spectral filtering. https://arxiv.org/abs/1606.09375,2017-02-05. |
| [9] | Wu Z H， Pan S R， Long G D，et al. Connecting the dots：Multivariate time series forecasting with graph neural networks∥Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining.New York，NY，USA：Association for Computing Machinery，2020：753-763. |
| [10] | Yu B， Yin H T， Zhu Z X. Spatio?temporal graph convolutional networks：A deep learning framework for traffic forecasting∥Proceedings of the 27th International Joint Conference on Artificial Intelli?gence. Palo Alto，CA，USA：AAAI Press，2018：3634-3640. |
| [11] | Li Y G， Yu R， Shahabi C，et al. Diffusion convolutional recurrent neural network：Data?driven traffic forecasting. https://arxiv.org/abs/1709. 04875，2018-02-22. |
| [12] | Pan Z Y， Liang Y X， Wang W F，et al. Urban traffic prediction from spatio?temporal data using deep meta learning∥Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. New York，NY，USA：Association for Computing Machinery，2019：1720-1730. |
| [13] | Zhao L， Song Y J， Zhang C，et al. T?GCN：A temporal graph convolutional network for traffic prediction. IEEE Transactions on Intelligent Transportation Systems，2020，21(9)：3848-3858. |
| [14] | Zheng C P， Fan X L， Wang C，et al. GMAN：A graph multi?attention network for traffic prediction. Proceedings of the AAAI Conference on Artificial Intelligence，2020，34(1)：1234-1241. |
| [15] | Guo S N， Lin Y F， Feng N，et al. Attention based spatial?temporal graph convolutional networks for traffic flow forecasting∥Proceedings of the AAAI Conference on Artificial Intelligence.Palo Alto，CA，USA：AAAI Press，2019：922-929. |
| [16] | Devlin J, Chang M W, Lee K,et al. BERT:Pre?training of deep bidirectional transformers for language understanding∥Proceedings of the 2019 Conference of the North. Minneapolis，MN，USA:Association for Computational Linguistics,2019:4171-4186. |
| [17] | Brown T B, Mann B, Ryder N,et al. Language models are few?shot learners∥Proceedings of the 34th International Conference on Neural Information Processing Systems.Red Hook，NY，USA:Curran Associates Inc.,2020:1877-1901. |
| [18] | Peters M E， Neumann M， Iyyer M，et al. Deep contextualized word representations∥Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics：Human Language Technologies. Volume 1：Long Papers. New Orleans，LA，USA：Association for Computational Linguistics，2018：2227-2237. |
| [19] | Sutskever I， Vinyals O， Le Quoc V. Sequence to sequence learning with neural networks. Advances in Neural Information Processing Systems，2014，27：3104-3112. |
| [20] | Dosovitskiy A， Beyer L， Kolesnikov A，et al. An image is worth 16×16 words：Transformers for image recognition at scale. https://arxiv.org/abs/2010.11929，2021-06-03. |
| [21] | Vaswani A， Shazeer N， Parmar N，et al. Attention is all you need∥Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook，NY，USA：Curran Associates Inc.，2017：6000-6010. |
| [22] | He K M， Chen X L， Xie S N，et al. Masked autoencoders are scalable vision learners∥2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans，LA，USA：IEEE，2022：15979-15988. |
| [23] | Franceschi L， Niepert M， Pontil M，et al. Learning discrete structures for graph neural networks∥Proceedings of the 36th International Conference on Machine Learning. New York，NY，USA：PMLR，2019：1972-1982. |
| [24] | Kipf T， Fetaya E， Wang K C，et al. Neural relational inference for interacting systems∥Proceedings of the 35th International Conference on Machine Learning. New York，NY，USA：PMLR，2018：2688-2697. |
| [25] | Li F X， Feng J， Yan H，et al. Dynamic graph convolutional recurrent network for traffic prediction：Benchmark and solution.ACM Transactions on Knowledge Discovery From Data，2023，17(1)：1-21. |
| [26] | Jang E， Gu S X， Poole B. Categorical reparameterization with gumbel?softmax. https://arxiv.org/abs/1611.01144，2017-08-05. |
| [27] | Maddison C J， Mnih A， Teh Y W. The concrete distribution：A continuous relaxation of discrete random variables. https://arxiv.org/abs/1611. 00712,2017-03-05. |
| [28] | Wu Z H， Pan S R， Long G D，et al. Graph wavenet for deep spatial?temporal graph modeling∥Proceedings of the 28th International Joint Conference on Artificial Intelligence. Macao，China：AAAI Press，2019：1907-1913. |
| [29] | Jiang R H， Wang Z N， Yong J W，et al. Spatio?temporal meta?graph learning for traffic forecasting. Proceedings of the AAAI Conference on Artificial Intelligence，2023，37(7)：8078-8086. |
| [30] | Chen S， Chen J， Bi J.Discrete graph structure learning for forecasting multiple time series：Report No：OSTI?1866733. Washington DC，USA：US Department of Energy Office of Scientific and Technical Information，2021. |
| [31] | Cao D F， Wang Y J， Duan J Y，et al. Spectral temporal graph neural network for multivariate time?series forecasting∥Proceedings of the 34th International Conference on Neural Information Processing Systems. Red Hook，NY，USA：Curran Associates Inc.，2020：17766-17778. |
| [32] | Bai L， Yao L N， Li C，et al. Adaptive graph convolutional recurrent network for traffic forecasting∥Proceedings of the 34th International Conference on Neural Information Processing Systems. Red Hook，NY，USA：Curran Associates Inc.，2020：17804-17815. |
), 闵帆1, 王欣1,3,4","risUrl_cn":"//www.sanmikaiseki.com/jns/CN/article/getTxtFile.do?fileType=Ris&id=1759","title_cn":"基于预训练机制的时空图气量预测模型","doi":"10.13232/j.cnki.jnju.2025.06.003","jieShuYe":"927","keywordList_en":["gas production prediction","pre?training mechanism","graph structure learning","spatio?temporal graph neural networks"],"endNoteUrl_cn":"//www.sanmikaiseki.com/jns/CN/article/getTxtFile.do?fileType=EndNote&id=1759","zhaiyao_en":"To address the limitations of traditional gas production prediction models in capturing long⁃term dependencies and modeling nonlinear，multi⁃scale spatial interactions among wells，this study proposes a pretraining⁃enhanced spatio⁃temporal graph model for gas production prediction. In the pre⁃training phase，a Transformer⁃based time series masked autoencoder is employed to extract deep temporal features from oilfield production data，generating segment⁃level representations with global context awareness and mitigating the inadequacy of conventional models in temporal feature extraction. During the prediction phase，a spatio⁃temporal feature fusion mechanism effectively integrates these temporal representations with spatial dependencies among gas wells captured by a dynamic graph structure learning module，thereby overcoming the performance bottlenecks of traditional approaches in scenarios involving long⁃term memory deficiency and incomplete predefined graph structures. Experimental results demonstrate that the proposed model achieves a mean absolute error of 0.156，yielding error reductions of 6.6%，36.3%，26.9% and 67.9% compared to baseline models，thus delivering significant performance improvements. This research establishes a novel deep learning paradigm for spatio⁃temporal data modeling in petroleum exploration and represents a promising step toward the engineering application of large⁃scale spatio⁃temporal models in the energy sector.
","bibtexUrl_en":"https://jns.nju.edu.cn/EN/article/getTxtFile.do?fileType=BibTeX&id=1759","abstractUrl_cn":"https://jns.nju.edu.cn/CN/10.13232/j.cnki.jnju.2025.06.003","zuoZheCn_L":"李继, 方宇, 闵帆, 王欣","juanUrl_cn":"https://jns.nju.edu.cn/CN/Y2025","lanMu_en":"","clcIndexList_cn":[{"code":"TE311","text":""}],"qiUrl_en":"https://jns.nju.edu.cn/EN/Y2025/V61/I6","zuoZhe_EN":"Ji Li1, Yu Fang1,2(), Fan Min1, Xin Wang1,3,4","risUrl_en":"https://jns.nju.edu.cn/EN/article/getTxtFile.do?fileType=Ris&id=1759","title_en":"A pretraining⁃enhanced spatio⁃temporal graph model for gas production prediction","hasPdf":"true"},"authorNotes_cn":["E⁃mail：fangyu@swpu.edu.cn
PDF(1196 KB)
PDF(1196 KB)
基于预训练机制的时空图气量预测模型
({{custom_author.role_cn}}),{{javascript:window.custom_author_cn_index++;}}A pretraining⁃enhanced spatio⁃temporal graph model for gas production prediction
({{custom_author.role_en}}),{{javascript:window.custom_author_en_index++;}}| {{custom_ref.label}} |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
/
| 〈 | 〉 |
