-
物理信息神经网络（PINNs）作为人工智能助力科学研究（AI for Science）求解偏微分方程（PDEs）的一种无网格化求解框架，近年来受到广泛关注。然而，传统PINNs存在局限性：一方面，PINNs网络结构使用单向信息传递的多层感知机（MLPs），难以有效聚焦序列数据中蕴含的关键特征，信息表征能力弱；另一方面，PINNs的损失函数为嵌入物理约束的二次罚函数，其未受约束而无限膨胀的惩罚因子影响模型训练寻优效率。为应对上述挑战，本文提出一种基于信息表征-损失优化改进的PINNs——allaPINNs，旨在增强模型关键特征提取和训练寻优能力，提升其求解PDEs数值解的准确性和泛化能力。在信息表征方面，allaPINNs引入高效线性注意力（LA）增强模型关键特征识别能力，同时降低权重动态加权的计算复杂度。在损失优化方面，allaPINNs通过引入增广拉格朗日（AL）函数重构目标损失函数，利用可学习的拉格朗日乘子和惩罚因子有效调控各损失残差项的相互作用。由Helmholtz、Black-Scholes、Burgers和非线性Schrödinger四个基准方程验证allaPINNs的有效性。结果表明，allaPINNs能够有效求解不同类型PDEs，并展现出卓越的数值解预测精度与泛化能力。相较于当前先进PINNs其预测精度提升一至两个数量级。Physics-informed neural networks (PINNs) have recently garnered significant attention as a meshless solution framework for solving partial differential equations (PDEs) in the context of AI-assisted scientific research (AI for Science). However, traditional PINNs exhibit certain limitations. On one hand, their network architecture, typically multilayer perceptrons (MLPs) with unidirectional information transfer, struggles to effectively capture key features embedded in sequential data, resulting in weak information characterization. On the other hand, the loss function of PINNs, a quadratic penalty function embedded with physical constraints, has an unconstrained and infinitely inflated penalty factor that affects the efficiency of the model’s training optimization search. To address these challenges, this paper proposes an improved PINN based on information representation and loss optimization, termed allaPINNs, which aims to enhance the model’s key feature extraction capability and training optimization search ability, thereby improving its accuracy and generalization for solving numerical solutions of PDEs. In terms of information characterization, allaPINNs introduces efficient linear attention (LA) to enhance the model’s ability to identify key features while reducing the computational complexity of dynamic weighting. In terms of loss optimization, allaPINNs reconstructs the objective loss function by introducing the augmented Lagrangian (AL) function, utilizing learnable Lagrangian multipliers and penalty factors to efficiently regulate the interaction of each loss residual term. The feasibility of allaPINNs is validated through four benchmark equations: Helmholtz, Black-Scholes, Burgers, and nonlinear Schrödinger. The results demonstrate that allaPINNs can effectively solve various PDEs of different complexities and exhibit excellent numerical solution prediction accuracy and generalization ability. Compared to the current state-of-the-art PINNs, the predictive accuracy is improved by one to two orders of magnitude.
-
Keywords:
- physics-informed neural networks /
- linear attention /
- augmented Lagrangian functions /
- solving partial differential equations
-
[1] Jin X, Cai S, Li H, Karniadakis G E 2021 J. Comput. Phys. 426109951
[2] Roul P, Goura V P 2020 J. Comput. Appl. Math. 363464
[3] Pu J C, Li J, Chen Y 2021 Chin. Phys. B 3060202
[4] Cuomo S, Di Cola V S, Giampaolo F, Rozza G, Raissi M, Piccialli F 2022 J. Sci. Comput. 9288
[5] Samaniego E, Anitescu C, Goswami S, Nguyen-Thanh V M, Guo H, Hamdia K, Zhuang X, Rabczuk T 2020 Comput. Methods Appl. Mech. Eng. 362112790
[6] Taylor C A, Hughes T J, Zarins C K 1998 Comput. Methods Appl. Mech. Eng. 158155
[7] Zhang Y 2009 Appl. Math. Comput. 215524
[8] Van Hoecke L, Boeye D, Gonzalez-Quiroga A, Patience G S, Perreault P 2023 Can. J. Chem. Eng. 101545
[9] Hasan F, Ali H, Arief H A 2025 Int. J. Appl. Comput. Math. 111
[10] Choo Y S, Choi N, Lee B C 2010 Appl. Math. Modell. 3414
[11] Lawrence N D, Montgomery J 2024 R. Soc. Open Sci. 11231130
[12] Si Z Z, Wang D L, Zhu B W, Ju Z T, Wang X P, Liu W, Malomed B A, Wang Y Y, Dai C Q 2024 Laser Photonics Rev. 182400097
[13] Fang Y, Han H B, Bo W B, Liu W, Wang B H, Wang Y Y, Dai C Q 2023 Opt. Lett. 48779
[14] Li N, Xu S, Sun Y, Chen Q 2025 Nonlinear Dyn. 113767
[15] Mouton L, Reiter F, Chen Y, Rebentrost P 2024 Phys. Rev. A 110022612
[16] Zhu M, Feng S, Lin Y, Lu L 2023 Comput. Methods Appl. Mech. Eng. 416116300
[17] Li X, Liu Z, Cui S, Luo C, Li C, Zhuang Z 2019 Comput. Methods Appl. Mech. Eng. 347735
[18] Wang S, Teng Y, Perdikaris P 2021 SIAM J. Sci. Comput. 43 A3055
[19] Chew A W Z, He R, Zhang L 2025 Arch. Comput. Methods Eng. 32399
[20] Bai J, Rabczuk T, Gupta A, Alzubaidi L, Gu Y 2023 Comput. Mech. 71543
[21] Son S, Lee H, Jeong D, Oh K Y, Sun K H 2023 Adv. Eng. Inf. 57102035
[22] Fang Z, Pan Y Q, Dai D, Zhang J B 2024 Acta Phys. Sin. 73145201(in Chinses) [方泽, 潘泳全, 戴栋, 张俊勃2024 73145201]
[23] Raissi M, Perdikaris P, Karniadakis G E 2019 J. Comput. Phys. 378686
[24] Hornik K 1991 Neural Netw. 4251
[25] Baydin A G, Pearlmutter B A, Radul A A, Siskind J M 2018 J. Mach. Learn. Res. 181
[26] De Ryck T, Mishra S 2024 Acta Numer. 33633
[27] Karniadakis G E, Kevrekidis I G, Lu L, Perdikaris P, Wang S, Yang L 2021 Nat. Rev. Phys. 3422
[28] Ren P, Rao C, Liu Y, Wang J X, Sun H 2022 Comput. Methods Appl. Mech. Eng. 389114399
[29] Lei L, He Y, Xing Z, Li Z, Zhou Y 2025 IEEE Trans. Ind. Inf. 215411
[30] Yuan B, Wang H, Heitor A, Chen X 2024 J. Comput. Phys. 515113284
[31] Wang Y, Sun J, Bai J, Anitescu C, Eshaghi M S, Zhuang X, Rabczuk T, Liu Y 2025 Comput. Methods Appl. Mech. Eng. 433117518
[32] Wang Y, Sun J, Bai J, Anitescu C, Eshaghi M S, Zhuang X, Rabczuk T, Liu Y 2025 Comput. Methods Appl. Mech. Eng. 433117518
[33] Jahani-Nasab M, Bijarchi M A 2024 Sci. Rep. 1423836
[34] Yu J, Lu L, Meng X, Karniadakis G E 2022 Comput. Methods Appl. Mech. Eng. 393114823
[35] Jiao Y, Lai Y, Lo Y, Wang Y, Yang Y 2024 Anal. Appl. 2257
[36] Yang A, Xu S, Liu H, Li N, Sun Y 2025 Nonlinear Dyn. 1131523
[37] Li Y, Zhou Z, Ying S 2022 J. Comput. Phys. 451110884
[38] Jacot A, Gabriel F, Hongler C 2018 Adv. Neural Inf. Process. Syst. 318570
[39] Xiang Z, Peng W, Liu X, Yao W 2022 Neurocomputing 49611
[40] Tancik M, Srinivasan P, Mildenhall B, Fridovich-Keil S, Raghavan N, Singhal U, Ramamoorthi R, Barron J, Ng R 2020 Adv. Neural Inf. Process. Syst. 337537
[41] Zhang Z, Wang Y, Tan S, Xia B, Luo Y 2025 Neurocomputing 625129429
[42] Zhang W, Li H, Tang L, Gu X, Wang L, Wang L 2022 Acta Geotech. 171367
[43] Zhang Z, Wang Q, Zhang Y, Shen T, Zhang W 2025 Sci. Rep. 1510523
[44] Cybenko G 1989 Math. Control Signals Syst. 2303
[45] Wang C, Ma C, Zhou J 2014 J. Global Optim. 5851
[46] Yi K, Zhang Q, Fan W, Wang S, Wang P, He H, An N, Lian D, Cao L, Niu Z 2023 Adv. Neural Inf. Process. Syst. 3676656
[47] Durstewitz D, Koppe G, Thurm M I 2023 Nat. Rev. Neurosci. 24693
[48] Chang G, Hu S, Huang H 2023 J. Supercomput. 796991
[49] Ocal H 2025 Arabian J. Sci. Eng. 501097
[50] Wu H C 2009 Eur. J. Oper. Res. 19649
[51] Curtis F E, Jiang H, Robinson D P 2015 Math. Program. 152201
[52] Sun D, Sun J, Zhang L 2008 Math. Program. 114349
[53] Kanzow C, Steck D 2019 Math. Program. 177425
[54] Rockafellar R T 2023 Math. Program. 198159
[55] Dampfhoffer M, Mesquida T, Valentian A, Anghel L 2023 IEEE Trans. Neural Networks Learn. Syst. 3511906
[56] Humbird K D, Peterson J L, McClarren R G 2018 IEEE Trans. Neural Networks Learn. Syst. 301286
[57] Zhang Z, Wang Q, Zhang Y, Shen T 2025 Digital Signal Process. 156104766
[58] Zhou P, Xie X, Lin Z, Yan S 2024 IEEE Trans. Pattern Anal. Mach. Intell. 466486
[59] Rather I H, Kumar S, Gandomi A H 2024 Artif. Intell. Rev. 57226
[60] Thulasidharan K, Priya N V, Monisha S, Senthilvelan M 2024 Phys. Lett. A 511129551
[61] Son H, Cho S W, Hwang H J 2023 Neurocomputing 548126424
[62] Song Y, Wang H, Yang H, Taccari M L, Chen X 2024 J. Comput. Phys. 501112781
计量
- 文章访问数: 65
- PDF下载量: 3
- 被引次数: 0
下载:
