[1]邵绪强,栗明宇,韩浩,等.深度学习方法在流场重建中的应用综述[J].智能系统学报,2026,21(1):2-18.[doi:10.11992/tis.202501017]
SHAO Xuqiang,LI Mingyu,HAN Hao,et al.Overview of the application of deep learning methods in flow field reconstruction[J].CAAI Transactions on Intelligent Systems,2026,21(1):2-18.[doi:10.11992/tis.202501017]
点击复制
SHAO Xuqiang,LI Mingyu,HAN Hao,et al.Overview of the application of deep learning methods in flow field reconstruction[J].CAAI Transactions on Intelligent Systems,2026,21(1):2-18.[doi:10.11992/tis.202501017]
深度学习方法在流场重建中的应用综述
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
21
期数:
2026年第1期
页码:
2-18
栏目:
综述
出版日期:
2026-03-05
- Title:
- Overview of the application of deep learning methods in flow field reconstruction
- Keywords:
- flow field reconstruction; deep learning; neural networks; computational fluid dynamics; numerical simulation; mode decomposition; super-resolution; data augmentation
- 分类号:
- TP391.9
- 摘要:
- 高分辨率流场数据具有非线性，数据量大的特点，无论用实验还是模拟方法都存在获取难度高的问题。流场重建技术能够充分利用流场的可观测信息挖掘不可观测信息，用稀疏观测的或低分辨的流场数据恢复出高分辨流场数据。深度学习方法得益于其强大的特征提取和非线性拟合能力，在流体力学问题中已经有了广泛的应用，其中，基于深度学习的流场重建方法拥有极高的研究潜力。本文对基于深度学习的流场重建方法进行了调研，分类阐述了不同视角下的流场重建问题的建模方式。详细归纳了模态重组类、局部-整体预测类和单元求解器类流场重建方法的研究进展和成果，并讨论了各种方法的优缺点。最后总结分析了基于深度学习的流场重建技术面临的挑战，并对未来的研究方向进行了展望。
- Abstract:
- High resolution flow field data has the characteristics of nonlinearity and large data volume, which makes it difficult to obtain through both experimental and simulation methods. Flow field reconstruction technology can fully utilize the observable information of the flow field to mine unobservable information, and recover high-resolution flow field data from sparse or low resolution flow field data. Deep learning methods have been widely applied in fluid mechanics problems due to their powerful feature extraction and nonlinear fitting capabilities. Among them, flow field reconstruction methods based on deep learning have high research potential. This article investigates deep learning based flow field reconstruction methods and categorizes modeling approaches for flow field reconstruction problems from different perspectives. This paper provides a detailed summary of the research progress and achievements in flow field reconstruction methods for modal recombination, local global prediction, and element solver, and discusses the advantages and disadvantages of each method. Finally, the challenges faced by deep learning based flow field reconstruction technology were summarized and analyzed, and future research directions were discussed.
- 参考文献/References:
-
[1] 周铸, 黄江涛, 黄勇, 等. CFD技术在航空工程领域的应用、挑战与发展[J]. 航空学报, 2017, 38(3): 20891-20891 ZHOU Zhu, HUANG Jiangtao, HUANG Yong, et al. Application, challenge and development of CFD technology in aviation engineering field[J]. Acta aeronautica et astronautica sinica, 2017, 38(3): 20891-20891
[2] ZHAO Qiang, LI Rui, CAO Kaifa, et al. Influence of building spatial patterns on wind environment and air pollution dispersion inside an industrial park based on CFD simulation[J]. Environmental monitoring and assessment, 2024, 196(5): 427
[3] CHEN Yuanqing, WANG Ding, FENG Dachuan, et al. Three-dimensional spatiotemporal wind field reconstruction based on LiDAR and multi-scale PINN[J]. Applied energy, 2025, 377: 124577
[4] 岳茂雄, 王如琴, 姚向红, 等. 高速聚焦纹影改进及应用[J]. 实验流体力学, 2013, 27(5): 88-93 YUE Maoxiong, WANG Ruqin, YAO Xianghong, et al. Improvement and application of high-speed focusing schlieren[J]. Journal of experiments in fluid mechanics, 2013, 27(5): 88-93
[5] 祁沛垚, 邓坚, 谭思超, 等. 基于PIV技术的低雷诺数下棒束通道流场研究[J]. 核动力工程, 2021(1): 18-22 QI Peiyao, DENG Jian, TAN Sichao, et al. Research on flow field in rod bundle channel under low Reynolds number using PIV technique[J]. Nuclear power engineering, 2021(1): 18-22
[6] 党冠麟, 刘世伟, 胡晓东, 等. 基于CPU/GPU异构系统架构的高超声速湍流直接数值模拟研究[J]. 数据与计算发展前沿, 2020, 2(1): 105-116 DANG Guanlin, LIU Shiwei, HU Xiaodong, et al. Direct numerical simulation of hypersonic turbulence based on CPU/GPU heterogeneous system architecture[J]. Frontiers of Data& Computing, 2020, 2(1): 105-116
[7] MANOHAR K, BRUNTON B W, KUTZ J N, et al. Data-driven sparse sensor placement for reconstruction: demonstrating the benefits of exploiting known patterns[J]. IEEE control systems magazine, 2018, 38(3): 63-86
[8] BOISSON J, DUBRULLE B. Three-dimensional magnetic field reconstruction in the VKS experiment through Galerkin transforms[J]. New journal of physics, 2011, 13(2): 023037
[9] ABRAHAMSON S, LONNES S. Uncertainty in calculating vorticity from 2D velocity fields using circulation and least-squares approaches[J]. Experiments in fluids, 1995, 20(1): 10-20
[10] 尹宇辉, 李浩然, 张宇飞, 等. 机器学习辅助湍流建模在分离流预测中的应用[J]. 空气动力学学报, 2021, 39(2): 23-32 YIN Yuhui, LI Haoran, ZHANG Yufei, et al. Application of turbulence modeling aided by machine learning in separated flow prediction[J]. Acta aerodynamica sinica, 2021, 39(2): 23-32
[11] OBIOLS-SALES O, VISHNU A, MALAYA N, et al. CFDNet: a deep learning-based accelerator for fluid simulations[C]//Proceedings of the 34th ACM International Conference on Supercomputing. New York: ACM, 2020: 1-12.
[12] ESFAHANIAN V, IZADI M J, BASHI H, et al. Aerodynamic shape optimization of gas turbines: a deep learning surrogate model approach[J]. Structural and multidisciplinary optimization, 2023, 67(1): 2
[13] XIA Chengwei, ZHANG Junjie, KERRIGAN E C, et al. Active flow control for bluff body drag reduction using reinforcement learning with partial measurements[J]. Journal of fluid mechanics, 2024, 981: A17
[14] HU Liwei, WANG Wenyong, XIANG Yu, et al. Flow field reconstructions with GANs based on radial basis functions[J]. IEEE transactions on aerospace and electronic systems, 2022, 58(4): 3460-3476
[15] PRUVOST J, LEGRAND J, LEGENTILHOMME P. Three-dimensional swirl flow velocity-field reconstruction using a neural network with radial basis functions[J]. Journal of fluids engineering, 2001, 123(4): 920-927
[16] BERKOOZ G, HOLMES P, LUMLEY J L. The proper orthogonal decomposition in the analysis of turbulent flows[J]. Annual review of fluid mechanics, 1993, 25: 539-575
[17] ARUN R, BAE H J, MCKEON B J. Towards real-time reconstruction of velocity fluctuations in turbulent channel flow[J]. Physical review fluids, 2023, 8(6): 064612
[18] SUN Shanxun, LIU Shi, LIU Jing, et al. Wind field reconstruction using inverse process with optimal sensor placement[J]. IEEE transactions on sustainable energy, 2019, 10(3): 1290-1299
[19] SUN Shanxun, LIU Shi, CHEN Minxin, et al. An optimized sensing arrangement in wind field reconstruction using CFD and POD[J]. IEEE transactions on sustainable energy, 2020, 11(4): 2449-2456
[20] EVERSON R, SIROVICH L. Karhunen-Loève procedure for gappy data[J]. Journal of the optical society of America A, 1995, 12(8): 1657-1664
[21] 李天一, MICHELE B, LUCA B, 等. Gappy POD方法重构湍流数据的研究[J]. 力学学报, 2021, 53(10): 2703-2711 LI Tianyi, MICHELE B, LUCA B, et al. Study on reconstruction of turbulence data by Gappy POD method[J]. Chinese journal of theoretical and applied mechanics, 2021, 53(10): 2703-2711
[22] CHINTA V K, LUHAR M. Statistically consistent resolvent-based reconstruction of turbulent channel flows from limited measurements[C]//12th International Symposium on Turbulence and Shear Flow Phenomena. Online: Begell House Inc., 2022.
[23] 袁昊, 寇家庆, 张伟伟. 流体力学预解分析方法研究进展[J]. 力学学报, 2024, 56(10): 2799-2814 YUAN Hao, KOU Jiaqing, ZHANG Weiwei. Research progress of resolvent analysis in fluid mechanics[J]. Chinese journal of theoretical and applied mechanics, 2024, 56(10): 2799-2814
[24] SNAIKI R, MIRFAKHAR S F. Multiresolution dynamic mode decomposition approach for wind pressure analysis and reconstruction around buildings[J]. Computer-aided civil and infrastructure engineering, 2024, 39(22): 3375-3391
[25] 寇家庆, 张伟伟. 动力学模态分解及其在流体力学中的应用[J]. 空气动力学学报, 2018, 36(2): 163-179 KOU Jiaqing, ZHANG Weiwei. Dynamic mode decomposition and its applications in fluid dynamics[J]. Acta aerodynamica sinica, 2018, 36(2): 163-179
[26] ZHANG Guangchao, LIU Shi. Reconstruction of unsteady wind field based on CFD and reduced-order model[J]. Mathematics, 2023, 11(10): 2223
[27] KOLDA T G, BADER B W. Tensor decompositions and applications[J]. SIAM review, 2009, 51(3): 455-500
[28] ABDULLAH A M S M, LU Chen, JAYARAMAN B, et al. Extreme learning machines as encoders for sparse reconstruction[J]. Fluids, 2018, 3(4): 88
[29] MURATA T, FUKAMI K, FUKAGATA K. Nonlinear mode decomposition with convolutional neural networks for fluid dynamics[J]. Journal of fluid mechanics, 2020, 882: A13
[30] FUKAMI K, NAKAMURA T, FUKAGATA K. Convolutional neural network based hierarchical autoencoder for nonlinear mode decomposition of fluid field data[J]. Physics of fluids, 2020, 32(9): 095110
[31] YU Jian, HESTHAVEN J S. Flowfield reconstruction method using artificial neural network[J]. AIAA journal, 2019, 57(2): 482-498
[32] DENG Zhiwen, CHEN Yujia, LIU Yingzheng, et al. Time-resolved turbulent velocity field reconstruction using a long short-term memory based artificial intelligence framework[J]. Physics of fluids, 2019, 31(7): 075108
[33] PENG Xingwen, LI Xingchen, CHEN Xiaoqian, et al. A hybrid deep learning framework for unsteady periodic flow field reconstruction based on frequency and residual learning[J]. Aerospace science and technology, 2023, 141: 108539
[34] MAULIK R, FUKAMI K, RAMACHANDRA N, et al. Probabilistic neural networks for fluid flow surrogate modeling and data recovery[J]. Physical review fluids, 2020, 5(10): 104401
[35] FUKAMI K, FUKAGATA K, TAIRA K. Assessment of supervised machine learning methods for fluid flows[J]. Theoretical and computational fluid dynamics, 2020, 34(4): 497-519
[36] ERICHSON N B, MATHELIN L, YAO Zhewei, et al. Shallow neural networks for fluid flow reconstruction with limited sensors[J]. Proceedings Mathematical, physical, and engineering sciences, 2020, 476(2238): 20200097
[37] LI Rui, SONG Baiyang, CHEN Yaoran, et al. Deep learning reconstruction of high-Reynolds-number turbulent flow field around a cylinder based on limited sensors[J]. Ocean engineering, 2024, 304: 117857
[38] FUKAMI K, FUKAGATA K, TAIRA K. Super-resolution reconstruction of turbulent flows with machine learning[J]. Journal of fluid mechanics, 2019, 870: 106-120
[39] DENG Zhiwen, HE Chuangxin, LIU Yingzheng, et al. Super-resolution reconstruction of turbulent velocity fields using a generative adversarial network-based artificial intelligence framework[J]. Physics of fluids, 2019, 31(12): 125111
[40] SARKAR R K, MAJUMDAR R, JADHAV V, et al. Redefining super-resolution: fine-mesh PDE predictions without classical simulations[EB/OL]. (2023-11-16)[2025-01-13]. https://arxiv.org/abs/2311.09740.
[41] SHEN Liming, DENG Liang, WANG Yueqing, et al. PCSAGAN: a physics-constrained generative network based on self-attention for high-fidelity flow field reconstruction[J]. Journal of visualization, 2024, 27(4): 661-676
[42] SHU Dule, LI Zijie, BARATI F A. A physics-informed diffusion model for high-fidelity flow field reconstruction[J]. Journal of computational physics, 2023, 478: 111972
[43] GUO Yanan, CAO Xiaoqun, ZHOU Mengge, et al. Enhancing high-resolution reconstruction of flow fields using physics-informed diffusion model with probability flow sampling[J]. Physics of fluids, 2024, 36(11): 115110
[44] SHAN Siming, WANG Pengkai, CHEN Song, et al. PiRD: physics-informed residual diffusion for flow field reconstruction[EB/OL]. (2024-04-12)[2025-01-13]. https://arxiv.org/abs/2404.08412.
[45] OBIOLS-SALES O, VISHNU A, MALAYA N, et al. NUNet: deep learning for non-uniform super-resolution of turbulent flows[EB/OL]. (2022-03-26)[2025-01-13]. https://arxiv.org/abs/2203.14154.
[46] OBIOLS-SALES O, VISHNU A, MALAYA N P, et al. SURFNet: super-resolution of turbulent flows with transfer learning using small datasets[C]//2021 30th International Conference on Parallel Architectures and Compilation Techniques. Piscataway: IEEE, 2021: 331-344.
[47] DE AVILA B P F, ECONOMON T D, KOLTER J Z. Combining differentiable PDE solvers and graph neural networks for fluid flow prediction[C]//International Conference on Machine Learning. [S. l. ]: JMLR, 2020.
[48] FUKAMI K, MAULIK R, RAMACHANDRA N, et al. Global field reconstruction from sparse sensors with Voronoi tessellation-assisted deep learning[J]. Nature machine intelligence, 2021, 3(11): 945-951
[49] NAKAMURA T, FUKAGATA K. Robust training approach of neural networks for fluid flow state estimations[J]. International journal of heat and fluid flow, 2022, 96: 108997
[50] DAW A, KARPATNE A, YEO K, et al. Source identification and field reconstruction of advection-diffusion process from sparse sensor measurements[C]//Machine Learning and the Physical Sciences Workshop. [S. l. ]: NeurIPS, 2023.
[51] GUNDERSEN K, OLEYNIK A, BLASER N, et al. Semi-conditional variational auto-encoder for flow reconstruction and uncertainty quantification from limited observations[J]. Physics of fluids, 2021, 33: 017119
[52] G?EMES A, SANMIGUEL VILA C, DISCETTI S. Super-resolution generative adversarial networks of randomly-seeded fields[J]. Nature machine intelligence, 2022, 4(12): 1165-1173
[53] LI Jinxing, LIU Tianyuan, WANG Yuqi, et al. Integrated graph deep learning framework for flow field reconstruction and performance prediction of turbomachinery[J]. Energy, 2022, 254: 124440.
[54] DUTH? G, ABDALLAH I, BARBER S, et al. Graph neural networks for aerodynamic flow reconstruction from sparse sensing[EB/OL]. (2023-01-09)[2025-01-13]. https://arxiv.org/abs/2301.03228.
[55] SUN Luning, WANG Jianxun. Physics-constrained Bayesian neural network for fluid flow reconstruction with sparse and noisy data[J]. Theoretical and applied mechanics letters, 2020, 10(3): 161-169
[56] MAI Jieai, LI Yang, LONG Lian, et al. Two-dimensional temperature field inversion of turbine blade based on physics-informed neural networks[J]. Physics of fluids, 2024, 36(3): 037114
[57] RAISSI M, YAZDANI A, KARNIADAKIS G E. Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations[J]. Science, 2020, 367(6481): 1026-1030
[58] CAI Shengze, WANG Zhicheng, FUEST F, et al. Flow over an espresso cup: inferring 3-D velocity and pressure fields from tomographic background oriented Schlieren via physics-informed neural networks[J]. Journal of fluid mechanics, 2021, 915: A102
[59] WANG Longyan, CHEN Meng, LUO Zhaohui, et al. Dynamic wake field reconstruction of wind turbine through Physics-Informed Neural Network and Sparse LiDAR data[J]. Energy, 2024, 291: 130401
[60] URCO J M, FERACO F, CHAU J L, et al. Augmented four-dimensional mesosphere and lower thermosphere wind field reconstruction via the physics-informed machine learning approach HYPER[J]. Journal of geophysical research: machine learning and computation, 2024, 1(3): e2024JH000162
[61] ZHU Yongzheng, CHEN Weizheng, DENG Jian, et al. Physics-informed neural networks for hidden boundary detection and flow field reconstruction[EB/OL]. (2025-03-31)[2025-01-13]. https://arxiv.org/abs/2503.24074.
[62] SANTOS J E, FOX Z R, MOHAN A, et al. Development of the Senseiver for efficient field reconstruction from sparse observations[J]. Nature machine intelligence, 2023, 5(11): 1317-1325
[63] MARCATO A, GUILTINAN E, VISWANATHAN H, et al. Journey over destination: dynamic sensor placement enhances generalization[J]. Machine learning: science and technology, 2024, 5(2): 025070
- 相似文献/References:
-
[1]张媛媛,霍静,杨婉琪,等.深度信念网络的二代身份证异构人脸核实算法[J].智能系统学报,2015,10(2):193.[doi:10.3969/j.issn.1673-4785.201405060]
ZHANG Yuanyuan,HUO Jing,YANG Wanqi,et al.A deep belief network-based heterogeneous face verification method for the second-generation identity card[J].CAAI Transactions on Intelligent Systems,2015,10():193.[doi:10.3969/j.issn.1673-4785.201405060]
[2]丁科,谭营.GPU通用计算及其在计算智能领域的应用[J].智能系统学报,2015,10(1):1.[doi:10.3969/j.issn.1673-4785.201403072]
DING Ke,TAN Ying.A review on general purpose computing on GPUs and its applications in computational intelligence[J].CAAI Transactions on Intelligent Systems,2015,10():1.[doi:10.3969/j.issn.1673-4785.201403072]
[3]马晓,张番栋,封举富.基于深度学习特征的稀疏表示的人脸识别方法[J].智能系统学报,2016,11(3):279.[doi:10.11992/tis.201603026]
MA Xiao,ZHANG Fandong,FENG Jufu.Sparse representation via deep learning features based face recognition method[J].CAAI Transactions on Intelligent Systems,2016,11():279.[doi:10.11992/tis.201603026]
[4]刘帅师,程曦,郭文燕,等.深度学习方法研究新进展[J].智能系统学报,2016,11(5):567.[doi:10.11992/tis.201511028]
LIU Shuaishi,CHENG Xi,GUO Wenyan,et al.Progress report on new research in deep learning[J].CAAI Transactions on Intelligent Systems,2016,11():567.[doi:10.11992/tis.201511028]
[5]马世龙,乌尼日其其格,李小平.大数据与深度学习综述[J].智能系统学报,2016,11(6):728.[doi:10.11992/tis.201611021]
MA Shilong,WUNIRI Qiqige,LI Xiaoping.Deep learning with big data: state of the art and development[J].CAAI Transactions on Intelligent Systems,2016,11():728.[doi:10.11992/tis.201611021]
[6]王亚杰,邱虹坤,吴燕燕,等.计算机博弈的研究与发展[J].智能系统学报,2016,11(6):788.[doi:10.11992/tis.201609006]
WANG Yajie,QIU Hongkun,WU Yanyan,et al.Research and development of computer games[J].CAAI Transactions on Intelligent Systems,2016,11():788.[doi:10.11992/tis.201609006]
[7]黄心汉.A3I:21世纪科技之光[J].智能系统学报,2016,11(6):835.[doi:10.11992/tis.201605022]
HUANG Xinhan.A3I: the star of science and technology for the 21st century[J].CAAI Transactions on Intelligent Systems,2016,11():835.[doi:10.11992/tis.201605022]
[8]宋婉茹,赵晴晴,陈昌红,等.行人重识别研究综述[J].智能系统学报,2017,12(6):770.[doi:10.11992/tis.201706084]
SONG Wanru,ZHAO Qingqing,CHEN Changhong,et al.Survey on pedestrian re-identification research[J].CAAI Transactions on Intelligent Systems,2017,12():770.[doi:10.11992/tis.201706084]
[9]杨梦铎,栾咏红,刘文军,等.基于自编码器的特征迁移算法[J].智能系统学报,2017,12(6):894.[doi:10.11992/tis.201706037]
YANG Mengduo,LUAN Yonghong,LIU Wenjun,et al.Feature transfer algorithm based on an auto-encoder[J].CAAI Transactions on Intelligent Systems,2017,12():894.[doi:10.11992/tis.201706037]
[10]王科俊,赵彦东,邢向磊.深度学习在无人驾驶汽车领域应用的研究进展[J].智能系统学报,2018,13(1):55.[doi:10.11992/tis.201609029]
WANG Kejun,ZHAO Yandong,XING Xianglei.Deep learning in driverless vehicles[J].CAAI Transactions on Intelligent Systems,2018,13():55.[doi:10.11992/tis.201609029]
备注/Memo
收稿日期:2025-1-25。
基金项目:国家重点研发计划项目（2021YFF0604000）.
作者简介:邵绪强，副教授，博士，主要研究方向为计算机物理动画、机器学习、可视化技术。以第一作者或通信作者发表学术论文40余篇，主持科研项目20余项。E-mail：shaoxuqiang@163.com。;栗明宇，硕士研究生，主要研究方向为深度学习、计算流体力学。E-mail：uu6666123@126.com。;韩浩，研究员，博士，主要研究方向为人工智能在化学污染和生物防控领域的应用。E-mail：thinkinghh@163.com。
通讯作者:韩浩. E-mail：thinkinghh@163.com
更新日期/Last Update:
2026-01-05