Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
184 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Graph Neural PDE Solvers with Conservation and Similarity-Equivariance (2405.16183v1)

Published 25 May 2024 in cs.LG, cs.AI, cs.CE, and physics.comp-ph

Abstract: Utilizing machine learning to address partial differential equations (PDEs) presents significant challenges due to the diversity of spatial domains and their corresponding state configurations, which complicates the task of encompassing all potential scenarios through data-driven methodologies alone. Moreover, there are legitimate concerns regarding the generalization and reliability of such approaches, as they often overlook inherent physical constraints. In response to these challenges, this study introduces a novel machine-learning architecture that is highly generalizable and adheres to conservation laws and physical symmetries, thereby ensuring greater reliability. The foundation of this architecture is graph neural networks (GNNs), which are adept at accommodating a variety of shapes and forms. Additionally, we explore the parallels between GNNs and traditional numerical solvers, facilitating a seamless integration of conservative principles and symmetries into machine learning models. Our findings from experiments demonstrate that the model's inclusion of physical laws significantly enhances its generalizability, i.e., no significant accuracy degradation for unseen spatial domains while other models degrade. The code is available at https://github.com/yellowshippo/fluxgnn-icml2024.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (39)
  1. Two-point step size gradient methods. IMA Journal of Numerical Analysis, 8(1):141–148, 1988.
  2. Relational inductive biases, deep learning, and graph networks. arXiv preprint arXiv:1806.01261, 2018.
  3. On the hyperparameters influencing a PINN’s generalization beyond the training domain. arXiv preprint arXiv:2302.07557, 2023.
  4. Message passing neural PDE solvers. In International Conference on Learning Representations, 2022. URL https://openreview.net/forum?id=vSix3HPYKSU.
  5. Learning to simulate and design for structural engineering. In International Conference on Machine Learning, pp. 1426–1436. PMLR, 2020.
  6. The finite volume method in computational fluid dynamics: an advanced introduction with OpenFOAM® and Matlab®. Springer, 2016.
  7. E (n)𝑛(n)( italic_n ) equivariant message passing simplicial networks. In International Conference on Machine Learning, pp. 9071–9081. PMLR, 2023.
  8. SE(3)-transformers: 3D roto-translation equivariant attention networks. Advances in Neural Information Processing Systems, 33, 2020.
  9. Neural message passing for quantum chemistry. In International Conference on Machine Learning, pp. 1263–1272. JMLR. org, 2017.
  10. Learning physical models that can respect conservation laws. Physica D: Nonlinear Phenomena, 457:133952, 2024. ISSN 0167-2789. doi: https://doi.org/10.1016/j.physd.2023.133952. URL https://www.sciencedirect.com/science/article/pii/S0167278923003068.
  11. Physics-embedded neural networks: Graph neural PDE solvers with mixed boundary conditions. In Oh, A. H., Agarwal, A., Belgrave, D., and Cho, K. (eds.), Advances in Neural Information Processing Systems, 2022. URL https://openreview.net/forum?id=B3TOg-YCtzo.
  12. Isometric transformation invariant and equivariant graph convolutional networks. In International Conference on Learning Representations, 2021. URL https://openreview.net/forum?id=FX0vR39SJ5q.
  13. Equivariant graph mechanics networks with constraints. In International Conference on Learning Representations, 2022. URL https://openreview.net/forum?id=SHbhHHfePhP.
  14. Jasak, H. Error analysis and estimation for the finite volume method with applications to fluid flows. 1996.
  15. Kannappan, P. Functional equations and inequalities with applications. Springer Science & Business Media, 2009.
  16. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014.
  17. Machine learning–accelerated computational fluid dynamics. Proceedings of the National Academy of Sciences, 118(21):e2101784118, 2021.
  18. Data-driven fluid simulations using regression forests. ACM Transactions on Graphics (TOG), 34(6):1–9, 2015.
  19. Learning skillful medium-range global weather forecasting. Science, pp.  eadi2336, 2023.
  20. Systems of conservation laws. Communications on Pure and Applied Mathematics, 13(2):217–237, 1960. doi: https://doi.org/10.1002/cpa.3160130205. URL https://onlinelibrary.wiley.com/doi/abs/10.1002/cpa.3160130205.
  21. LeVeque, R. J. Numerical methods for conservation laws, volume 214. Springer, 1992.
  22. Rapidly encoding generalizable dynamics in a euclidean symmetric neural network. Extreme Mechanics Letters, 58:101925, 2023.
  23. Reynolds averaged turbulence modelling using deep neural networks with embedded invariance. Journal of Fluid Mechanics, 807:155–166, 2016.
  24. Deep energy-based modeling of discrete-time physics. Advances in Neural Information Processing Systems, 33:13100–13111, 2020.
  25. Vc-pinn: Variable coefficient physics-informed neural network for forward and inverse problems of pdes with variable coefficient. Physica D: Nonlinear Phenomena, 456:133945, 2023.
  26. Pytorch: An imperative style, high-performance deep learning library. In Wallach, H., Larochelle, H., Beygelzimer, A., d'Alché-Buc, F., Fox, E., and Garnett, R. (eds.), Advances in Neural Information Processing Systems, pp.  8024–8035. Curran Associates, Inc., 2019.
  27. Learning mesh-based simulation with graph networks. In International Conference on Learning Representations, 2021. URL https://openreview.net/forum?id=roNqYL0_XP.
  28. A graph convolutional autoencoder approach to model order reduction for parametrized pdes. Journal of Computational Physics, pp.  112762, 2024.
  29. A fast matrix-free elasto-plastic solver for predicting residual stresses in additive manufacturing. Computer-Aided Design, 123:102829, 2020.
  30. Neural conservation laws: A divergence-free perspective. Advances in Neural Information Processing Systems, 35:38075–38088, 2022.
  31. Learning to simulate complex physics with graph networks. In International conference on machine learning, pp. 8459–8468. PMLR, 2020.
  32. E (n) equivariant graph neural networks. In International conference on machine learning, pp. 9323–9332. PMLR, 2021.
  33. Scale-equivariant steerable networks. In International Conference on Learning Representations, 2020. URL https://openreview.net/forum?id=HJgpugrKPS.
  34. Tensor field networks: Rotation-and translation-equivariant neural networks for 3d point clouds. arXiv preprint arXiv:1802.08219, 2018.
  35. Computational fluid dynamics. The finite volume method, pp.  1–26, 1995.
  36. Incorporating symmetry into deep dynamics models for improved generalization. In International Conference on Learning Representations, 2021. URL https://openreview.net/forum?id=wta_8Hx2KD.
  37. Deep scale-spaces: Equivariance over scale. Advances in Neural Information Processing Systems, 32, 2019.
  38. Scale-equivariant unet for histopathology image segmentation. arXiv preprint arXiv:2304.04595, 2023.
  39. Deep sets. Advances in Neural Information Processing Systems, 30, 2017.
Citations (2)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets