物理引导的深度学习研究综述：进展、挑战和展望

doi:10.3778/j.issn.1673-9418.2407056

摘要/Abstract

摘要： 尽管深度学习在处理非线性高维问题时表现出强大的能力，但在复杂科学与工程问题中仍面临诸多挑战，如高昂的计算成本、大量的数据需求、难以解释的黑盒特性，缺乏对物理规律的建模能力等。为此，近年来涌现了一种新的框架——物理引导深度学习，通过将领域内的物理知识融入深度学习模型的构建和训练过程中，旨在增强模型的性能、可解释性及其物理一致性。对国内外关于物理引导深度学习的相关工作进行了全面梳理与分析。介绍了物理引导深度学习框架的主要动机与理论基础。对物理信息组合与物理信息融合两种模式进行了详细讨论，总结了各方法的特点、局限性与应用场景。分析了物理引导深度学习在多个领域应用中的表现，并从计算复杂性与优化收敛问题、控制方程偏离问题、观测数据依赖问题与知识融合困难问题四个方面探讨了该框架目前面临的挑战，并基于此展望该领域未来的发展方向，以期为研究者提供借鉴思路及多维度视角。

关键词: 科学范式, 物理引导, 深度学习, 模型融合, 控制方程

Abstract: Although deep learning has significant achievements in addressing nonlinear and high-dimensional problems, it faces challenges in complex scientific and engineering domains (such as high computational costs and data requirements, the difficulties in interpreting its black-box nature, and the lack of capabilities for following the physical laws). Therefore, a novel framework called physics-guided deep learning has emerged which enhances the performance, explainability, and physical consistency of deep learning by integrating domain-specific physical knowledge into the construction and training process of deep learning models. This paper reviews and analyzes the researches (e.g., methodologies, applications, etc.) on physics-guided deep learning thoroughly. Firstly, the main motivations and theoretical foundations of the physics-guided deep learning are introduced. Secondly, a detailed discussion is conducted on the two modes: the combination of physical information with deep learning and the fusion of physical information with deep learning. The characteristics, limitations and application scenarios of the two modes are summarized and discussed. Finally, the performance of physics-guided deep learning on various applications is analyzed. Furthermore, the challenges of the physics-guided deep learning are discussed from four perspectives: computational complexity and convergence, biases while involving control equations, dependence on observational data, and difficulties in knowledge fusion, based on which, an outlook for the future direction of this domain is provided. This paper strives for providing research reference and multidimensional perspectives of physics-guided deep learning for the researchers.

Key words: scientific paradigm, physics-guided, deep learning, model fusion, governing equations

陈冲, 朱啸宇, 王芳, 许雅倩, 张伟. 物理引导的深度学习研究综述：进展、挑战和展望[J]. 计算机科学与探索, 2025, 19(2): 277-294.

CHEN Chong, ZHU Xiaoyu, WANG Fang, XU Yaqian, ZHANG Wei. Comprehensive Review of Physics-Guided Deep Learning: Advancements, Challenges, and Perspectives[J]. Journal of Frontiers of Computer Science and Technology, 2025, 19(2): 277-294.

参考文献

[1] 邓仲华, 李志芳. 科学研究范式的演化: 大数据时代的科学研究第四范式[J]. 情报资料工作, 2013, 34(4): 19-23.
DENG Z H, LI Z F. The evolution of scientific research paradigm: the fourth paradigm of scientific research in the era of big data[J]. Information and Documentation Services, 2013, 34(4): 19-23.
[2] TOLLE K M, TANSLEY D S W, HEY A J G. The fourth paradigm: data-intensive scientific discovery[J]. Proceedings of the IEEE, 2011, 99(8): 1334-1337.
[3] 曾和义. 网格质量对数值模拟的影响[C]//中国核学会2009年学术年会论文集. 北京: 原子能出版社, 2009: 235-239.
ZENG H Y. On the influences of mesh quality upon numerical simulation[C]//Proceedings of the 2009 Academic Annual Conference of the Chinese Nuclear Society. Beijing:Atomic Energy Press, 2009: 235-239.
[4] 余凯, 贾磊, 陈雨强, 等. 深度学习的昨天、今天和明天[J]. 计算机研究与发展, 2013, 50(9): 1799-1804.
YU K, JIA L, CHEN Y Q, et al. Deep learning: yesterday, today, and tomorrow[J]. Journal of Computer Research and Development, 2013, 50(9): 1799-1804.
[5] HINTON G E, SALAKHUTDINOV R R. Reducing the dimensionality of data with neural networks[J]. Science, 2006, 313(5786): 504-507.
[6] LEWICKI G,MARINO G. Approximation by superpositions of a sigmoidal function[J]. Zeitschrift Fur Analysis Und Ihre Anwendungen, 2003, 22(2): 463-470.
[7] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. Image- Net classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90.
[8] WAN Z Y, VLACHAS P, KOUMOUTSAKOS P, et al. Data- assisted reduced-order modeling of extreme events in complex dynamical systems[J]. PLoS One, 2018, 13(5): e0197704.
[9] 牛程程, 李少波, 胡建军, 等. 机器学习在材料信息学中的应用综述[J]. 材料导报, 2020, 34(23): 23100-23108.
NIU C C, LI S B, HU J J, et al. Application of machine learning in material informatics: a survey[J]. Materials Reports, 2020, 34(23): 23100-23108.
[10] MAXWELL R M, CONDON L E, MELCHIOR P. A physics- informed, machine learning emulator of a 2D surface water model: what temporal networks and simulation-based inference can help us learn about hydrologic processes[J]. Water, 2021, 13(24): 3633.
[11] CHOUDHARY A, LINDNER J F, HOLLIDAY E G, et al. Physics-enhanced neural networks learn order and chaos[J]. Physical Review E, 2020, 101(6): 062207.
[12] COHEN T S, GEIGER M, KOEHLER J, et al. Spherical CNNs[EB/OL]. [2024-07-10]. https://arxiv.org/abs/1801.10130.
[13] THOMPSON M L, KRAMER M A. Modeling chemical processes using prior knowledge and neural networks[J]. AIChE Journal, 1994, 40(8): 1328-1340.
[14] RAISSI M, PERDIKARIS P, KARNIADAKIS G E. Physics informed deep learning (part II): data-driven discovery of nonlinear partial differential equations[EB/OL]. [2024-07-10]. https://arxiv.org/abs/1711.10566.
[15] RAISSI M, PERDIKARIS P, KARNIADAKIS G E. Physics informed deep learning (part I): data-driven solutions of nonlinear partial differential equations[EB/OL]. [2024-07-10]. https://arxiv.org/abs/1711.10561.
[16] WILLARD J, JIA X W, XU S M, et al. Integrating scientific knowledge with machine learning for engineering and environmental systems[J]. ACM Computing Surveys, 2022, 55(4): 1-37.
[17] VON RUEDEN L, MAYER S, BECKH K, et al. Informed machine learning — a taxonomy and survey of integrating prior knowledge into learning systems[J]. IEEE Transactions on Knowledge and Data Engineering, 2023, 35(1): 614-633.
[18] KARPATNE A, ATLURI G, FAGHMOUS J H, et al. Theory- guided data science: a new paradigm for scientific discovery from data[J]. IEEE Transactions on Knowledge and Data Engineering, 2017, 29(10): 2318-2331.
[19] KARNIADAKIS G E, KEVREKIDIS I G, LU L, et al. Physics-informed machine learning[J]. Nature Reviews Physics, 2021, 3: 422-440.
[20] DE PAUL ADOMBI A V, CHESNAUX R, BOUCHER M A. Review: theory-guided machine learning applied to hydrogeology: state of the art, opportunities and future challenges[J]. Hydrogeology Journal, 2021, 29(8): 2671-2683.
[21] YU Y, YAO H P, LIU Y M. Structural dynamics simulation using a novel physics-guided machine learning method[J]. Engineering Applications of Artificial Intelligence, 2020, 96: 103947.
[22] YAO K, HERR J E, TOTH D W, et al. The TensorMol-0.1 model chemistry: a neural network augmented with long-range physics[J]. Chemical Science, 2018, 9(8): 2261-2269.
[23] HAUTIER G, FISCHER C C, JAIN A, et al. Finding nature?s missing ternary oxide compounds using machine learning and density functional theory[J]. Chemistry of Materials, 2010, 22(12): 3762-3767.
[24] XU R, ZHANG D X, RONG M, et al. Weak form theory-guided neural network (TgNN-wf) for deep learning of subsurface single- and two-phase flow[J]. Journal of Computational Physics, 2021, 436: 110318.
[25] XIE K, LIU P, ZHANG J Y, et al. Physics-guided deep learning for rainfall-runoff modeling by considering extreme events and monotonic relationships[J]. Journal of Hydrology, 2021, 603: 127043.
[26] WANG N Z, ZHANG D X, CHANG H B, et al. Deep learning of subsurface flow via theory-guided neural network[J]. Journal of Hydrology, 2020, 584: 124700.
[27] WANG N Z, CHANG H B, ZHANG D X. Deep-learning-based inverse modeling approaches: a subsurface flow example[J]. Journal of Geophysical Research: Solid Earth, 2021, 126(2): e2020JB020549.
[28] READ J S, JIA X W, WILLARD J, et al. Process-guided deep learning predictions of lake water temperature[J]. Water Resources Research, 2019, 55(11): 9173-9190.
[29] DAW A, KARPATNE A, WATKINS W D, et al. Physics-guided neural networks (PGNN): an application in lake temperature modeling[M]//Knowledge-guided machine learning. Boca Raton: Chapman and Hall/CRC, 2022: 353-372.
[30] JIA X W, ZWART J, SADLER J, et al. Physics-guided recurrent graph model for predicting flow and temperature in river networks[C]//Proceedings of the 2021 SIAM International Conference on Data Mining. Philadelphia: SIAM, 2021: 612-620.
[31] KHANDELWAL A, MITHAL V, KUMAR V. Post classification label refinement using implicit ordering constraint among data instances[C]//Proceedings of the 2015 IEEE International Conference on Data Mining. Piscataway: IEEE, 2015: 799-804.
[32] YIN X F, HUANG Z, LIU Y. Bridge damage identification under the moving vehicle loads based on the method of physics-guided deep neural networks[J]. Mechanical Systems and Signal Processing, 2023, 190: 110123.
[33] KASHEFI A, REMPE D, GUIBAS L J. A point-cloud deep learning framework for prediction of fluid flow fields on irregular geometries[J]. Physics of Fluids, 2021, 33(2): 027104.
[34] KHANDELWAL A, KARPATNE A, MARLIER M E, et al. An approach for global monitoring of surface water extent variations in reservoirs using MODIS data[J]. Remote Sensing of Environment, 2017, 202: 113-128.
[35] CHEN Y T, HUANG D, ZHANG D X, et al. Theory-guided hard constraint projection (HCP): a knowledge-based data-driven scientific machine learning method[J]. Journal of Computational Physics, 2021, 445: 110624.
[36] WADA K, SUZUKI K, YONEKURA K. Physics-guided training of GAN to improve accuracy in airfoil design synthesis[J]. Computer Methods in Applied Mechanics and Engineering, 2024, 421: 116746.
[37] JOHNSON R, FILALI BOUBRAHIMI S, BAHRI O, et al. Combining empirical and physics-based models for solar wind prediction[J]. Universe, 2024, 10(5): 191.
[38] LU W J, WANG Y, ZHANG M Q, et al. Physics guided neural network: remaining useful life prediction of rolling bearings using long short-term memory network through dynamic weighting of degradation process[J]. Engineering Applications of Artificial Intelligence, 2024, 127: 107350.
[39] MAGRI L, DOAN N. First-principles machine learning modelling of COVID-19[EB/OL]. [2024-07-10]. https://arxiv. org/abs/2004.09478.
[40] XIE Y C, CHI H H, WANG Y H, et al. Physics-specialized neural network with hard constraints for solving multi-material diffusion problems[J]. Computer Methods in Applied Mechanics and Engineering, 2024, 430: 117223.
[41] LI X A, DENG J X, WU J R, et al. Physical informed neural networks with soft and hard boundary constraints for solving advection-diffusion equations using Fourier expansions[J]. Computers & Mathematics with Applications, 2024, 159: 60-75.
[42] MCCULLOCH W S, PITTS W. A logical calculus of the ideas immanent in nervous activity[J]. Bulletin of Mathematical Biology, 1990, 52(1/2): 99-115.
[43] GREFF K, SRIVASTAVA R K, KOUTNíK J, et al. LSTM: a search space odyssey[J]. IEEE Transactions on Neural Networks and Learning Systems, 2016, 28(10): 2222-2232.
[44] JIA X W, WILLARD J, KARPATNE A, et al. Physics-guided machine learning for scientific discovery: an application in simulating lake temperature profiles[J]. ACM/IMS Transactions on Data Science, 2021, 2(3): 1-26.
[45] ZHONG L J, LEI H M, GAO B. Developing a physics-informed deep learning model to simulate runoff response to climate change in alpine catchments[J]. Water Resources Research, 2023, 59(6): e2022WR034118.
[46] GAO M S, YU J, YANG Z F, et al. A physics-guided graph convolution neural network for optimal power flow[J]. IEEE Transactions on Power Systems, 2023, 39(1): 380-390.
[47] JIANG C, KASHINATH K, PRABHAT, et al. Enforcing physical constraints in CNNs through differentiable PDE layer[C]//Proceedings of the ICLR 2020 Workshop on Integration of Deep Neural Models and Differential Equations, 2020.
[48] HARADA K, RAJARAM D, MAVRIS D N. Application of transfer learning for multi-fidelity regression using physics-informed neural network on an airfoil[C]//Proceedings of the AIAA SCITECH 2022 Forum. Reston: AIAA, 2022: 0386.
[49] CHEN S Y, KALANAT N, XIE Y Q, et al. Physics-guided machine learning from simulated data with different physical parameters[J]. Knowledge and Information Systems, 2023, 65(8): 3223-3250.
[50] PAWAR S, SAN O, AKSOYLU B, et al. Physics guided machine learning using simplified theories[J]. Physics of Fluids, 2021, 33(1): 011701.
[51] HANACHI H, YU W N, KIM I Y, et al. Hybrid data-driven physics-based model fusion framework for tool wear prediction[J]. The International Journal of Advanced Manufacturing Technology, 2019, 101(9): 2861-2872.
[52] XU X B, LIU C. Physics-guided deep learning for damage detection in CFRP composite structures[J]. Composite Structures, 2024, 331: 117889.
[53] CHEN C, ZHANG H, SHI W X, et al. A novel paradigm for integrating physics-based numerical and machine learning models: a case study of eco-hydrological model[J]. Environmental Modelling & Software, 2023, 163: 105669.
[54] CHO K, KIM Y. Improving streamflow prediction in the WRF-Hydro model with LSTM networks[J]. Journal of Hydrology, 2022, 605: 127297.
[55] LI W T, PAN B X, XIA J J, et al. Convolutional neural network-based statistical post-processing of ensemble precipitation forecasts[J]. Journal of Hydrology, 2022, 605: 127301.
[56] JEON J, LEE J, VINUESA R, et al. Residual-based physics-informed transfer learning: a hybrid method for accelerating long-term CFD simulations via deep learning[J]. International Journal of Heat and Mass Transfer, 2024, 220: 124900.
[57] BISWAS R, SEN M K, DAS V, et al. Prestack and poststack inversion using a physics-guided convolutional neural network[J]. Interpretation, 2019, 7(3): SE161-SE174.
[58] WANG N Z, CHANG H B, ZHANG D X. Theory-guided auto-encoder for surrogate construction and inverse modeling[J]. Computer Methods in Applied Mechanics and Engineering, 2021, 385: 114037.
[59] KALAJAHI A P, PEREZ-RAYA I, D’SOUZA R M. Physics informed deep neural net inverse modeling for estimating model parameters in permeable porous media flows[J]. Journal of Fluids Engineering, 2022, 144(6): 061102.
[60] XU K L, TARTAKOVSKY A M, BURGHARDT J, et al. Learning viscoelasticity models from indirect data using deep neural networks[J]. Computer Methods in Applied Mechanics and Engineering, 2021, 387: 114124.
[61] XU K L, DARVE E. Physics constrained learning for data-driven inverse modeling from sparse observations[J]. Journal of Computational Physics, 2022, 453: 110938.
[62] WANG S F, WANG H W, PERDIKARIS P. Learning the solution operator of parametric partial differential equations with physics-informed DeepONets[J]. Science Advances, 2021, 7(40): eabi8605.
[63] DUVENAUD D, MACLAURIN D, AGUILERA-IPARAGUIRRE J, et al. Convolutional networks on graphs for learning molecular fingerprints[C]//Proceedings of the 29th International Conference on Neural Information Processing Systems - Volume 2. New York: ACM, 2015: 2224-2232.
[64] WANG S F, YU X L, PERDIKARIS P. When and why PINNs fail to train: a neural tangent kernel perspective[J]. Journal of Computational Physics, 2022, 449: 110768.
[65] RAISSI M, PERDIKARIS P, KARNIADAKIS G E. Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations[J]. Journal of Computational Physics, 2019, 378: 686-707.
[66] WANG S F, TENG Y J, PERDIKARIS P. Understanding and mitigating gradient flow pathologies in physics-informed neural networks[J]. SIAM Journal on Scientific Computing, 2021, 43(5): A3055-A3081.
[67] KE J H, MA J H, YIN X Y. Simulation and application of COVID-19 compartment model using physic-informed neural network[EB/OL]. [2024-07-10]. https://arxiv.org/abs/2208. 02433.
[68] DANDEKAR R, BARBASTATHIS G Quantifying the effect of quarantine control in Covid-19 infectious spread using machine learning[J]. medRxiv, 2020, 6: 1-13.
[69] HONG T Z, LEE S H. Integrating physics-based models with sensor data: an inverse modeling approach[J]. Building and Environment, 2019, 154: 23-31.
[70] BANDAI T, GHEZZEHEI T A. Physics-informed neural networks with monotonicity constraints for Richardson- Richards equation: estimation of constitutive relationships and soil water flux density from volumetric water content measurements[J]. Water Resources Research, 2021, 57(2): e2020WR027642.
[71] WANG S F, LI B W, CHEN Y H, et al. PirateNets: physics-informed deep learning with residual adaptive networks[EB/OL]. [2024-08-10]. https://arxiv.org/abs/2402.00326.
[72] RAISSI M, KARNIADAKIS G E. Hidden physics models: machine learning of nonlinear partial differential equations[J]. Journal of Computational Physics, 2018, 357: 125-141.
[73] GAUDING M, BODE M. Using physics-informed enhanced super-resolution generative adversarial networks to reconstruct mixture fraction statistics of turbulent jet flows[C]//Proceedings of the 35th International ISC High Performance Conference. Cham: Springer, 2021:138-153.
[74] SUN L N, GAO H, PAN S W, et al. Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data[J]. Computer Methods in Applied Mechanics and Engineering, 2020, 361: 112732.
[75] PANG G F, LU L, KARNIADAKIS G E. fPINNs: fractional physics-informed neural networks[J]. SIAM Journal on Scientific Computing, 2019, 41(4): A2603-A2626.
[76] LU L, PESTOURIE R, YAO W J, et al. Physics-informed neural networks with hard constraints for inverse design[J]. SIAM Journal on Scientific Computing, 2021, 43(6): B1105- B1132.
[77] ZHANG D K, LU L, GUO L, et al. Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems[J]. Journal of Computational Physics, 2019, 397: 108850.
[78] ZHANG J, MA W J. Data-driven discovery of governing equations for fluid dynamics based on molecular simulation[J]. Journal of Fluid Mechanics, 2020, 892: A5.
[79] RIEGER L, SINGH C, MURDOCH W, et al. Interpretations are useful: penalizing explanations to align neural networks with prior knowledge[C]//Proceedings of the 37th International Conference on Machine Learning. New York: ACM, 2020: 8116-8126.
[80] ZHANG Q S, ZHU S C. Visual interpretability for deep learning: a survey[J]. Frontiers of Information Technology & Electronic Engineering, 2018, 19(1): 27-39.
[81] GUERRERO-GóMEZ-OLMEDO R, SALMERON J L, KUCHKOVSKY C. LRP-based path relevances for global explanation of deep architectures[J]. Neurocomputing, 2020, 381: 252-260.
[82] SUN J M, BINDER A. Generalized PatternAttribution for neural networks with sigmoid activations[C]//Proceedings of the 2019 International Joint Conference on Neural Networks. Piscataway: IEEE, 2019: 1-9.
[83] IRRGANG C, BOERS N, SONNEWALD M, et al. Towards neural earth system modelling by integrating artificial intelligence in earth system science[J]. Nature Machine Intelligence, 2021, 3: 667-674.
[84] ZHANG S X, ZHANG J L, YANG L X, et al. Physics guided deep learning-based model for short-term origin–destination demand prediction in urban rail transit systems under pandemic[J]. Engineering, 2024, 41: 276-296.
[85] STOUDENMIRE E, SCHWAB D. Supervised learning with quantum-inspired tensor networks[C]//Advances in Neural Information Processing Systems 28, 2016: 4806-4814.
[86] SMITH G D. Numerical solution of partial differential equations: finite difference methods[M]. Oxford: Oxford University Press, 1985.
[87] GREENFELD D, GALUN M, KIMMEL R, et al. Learning to optimize multigrid PDE solvers[C]//Proceedings of the 36th International Conference on Machine Learning. San Diego: JMLR, 2019.
[88] 曹富军, 郭晓斌, 高飞, 等. 求解偏微分方程的卷积迭代方法[J]. 山西大学学报(自然科学版), 2023, 46(2): 293-303.
CAO F J, GUO X B, GAO F, et al. The convolutional iterative method for solving partial differential equation[J]. Journal of Shanxi University (Natural Science Edition), 2023, 46(2): 293-303.
[89] HOLL P, KOLTUN V, UM K, et al. phiflow: a differentiable PDE solving framework for deep learning via physical simulations[C]//Proceedings of the NeurIPS Workshop. Cambridge: MIT Press, 2020.
[90] SIRIGNANO J, SPILIOPOULOS K. DGM: a deep learning algorithm for solving partial differential equations[J]. Journal of Computational Physics, 2018, 375: 1339-1364.
[91] DISSANAYAKE M W M G, PHAN-THIEN N. Neural-network-based approximations for solving partial differential equations[J]. Communications in Numerical Methods in Engineering, 1994, 10(3): 195-201.
[92] LONG Z C, LU Y P, MA X Z, et al. PDE-Net: learning PDEs from data[C]//Proceedings of the 35th International Conference on Machine Learning. New York: PMLR, 2018: 3208-3216.
[93] LU L, MENG X H, MAO Z P, et al. DeepXDE: a deep learning library for solving differential equations[J]. SIAM Review, 2021, 63(1): 208-228.
[94] RUDY S H, BRUNTON S L, PROCTOR J L, et al. Data-driven discovery of partial differential equations[J]. Science Advances, 2017, 3(4): e1602614.
[95] LUAN L L, LIU Y, SUN H. Distilling governing laws and source input for dynamical systems from videos[C]//Proceedings of the 31st International Joint Conference on Artificial Intelligence. Vienna: IJCAI Organization, 2022: 3898-3904.
[96] CHEN B Y, HUANG K, RAGHUPATHI S, et al. Automated discovery of fundamental variables hidden in experimental data[J]. Nature Computational Science, 2022, 2(7): 433-442.
[97] RAISSI M, PERDIKARIS P, KARNIADAKIS G E. Machine learning of linear differential equations using Gaussian processes[J]. Journal of Computational Physics, 2017, 348: 683-693.
[98] CHO Y, KIM S, LI P P, et al. Physics-guided reinforcement learning for 3D molecular structures[C]//Proceedings of the Workshop at the 33rd Conference on Neural Information Processing Systems. Cambridge: MIT Press, 2019.
[99] FRAEHR N, WANG Q J, WU W Y, et al. Assessment of surrogate models for flood inundation: the physics-guided LSG model vs. state-of-the-art machine learning models[J]. Water Research, 2024, 252: 121202.
[100] BASEMAN E, DEBARDELEBEN N, BLANCHARD S, et al. Physics-informed machine learning for DRAM error modeling[C]//Proceedings of the 2018 IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems. Piscataway: IEEE, 2018: 1-6.
[101] DOAN N A K, POLIFKE W, MAGRI L. Physics-informed echo state networks for chaotic systems forecasting[C]//Proceedings of the 19th Annual International Conference on Computational Science. Cham: Springer, 2019: 192-198.
[102] FENG Y Y, SONG X Z, YUAN W, et al. Physics-informed deep learning cascade loss model[J]. Aerospace Science and Technology, 2023, 134: 108165.
[103] HINTON G E. How neural networks learn from experience[J]. Scientific American, 1992, 267(3): 144-151.
[104] 查文舒, 李道伦, 沈路航, 等. 基于神经网络的偏微分方程求解方法研究综述[J]. 力学学报, 2022, 54(3): 543-556.
ZHA W S, LI D L, SHEN L H, et al. Review of neural network-based methods for solving partial differential equations[J]. Chinese Journal of Theoretical and Applied Mechanics, 2022, 54(3): 543-556.