Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
169 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

An Implicit GNN Solver for Poisson-like problems (2302.10891v3)

Published 6 Feb 2023 in cs.LG, cs.AI, and math.AP

Abstract: This paper presents $\Psi$-GNN, a novel Graph Neural Network (GNN) approach for solving the ubiquitous Poisson PDE problems with mixed boundary conditions. By leveraging the Implicit Layer Theory, $\Psi$-GNN models an "infinitely" deep network, thus avoiding the empirical tuning of the number of required Message Passing layers to attain the solution. Its original architecture explicitly takes into account the boundary conditions, a critical prerequisite for physical applications, and is able to adapt to any initially provided solution. $\Psi$-GNN is trained using a "physics-informed" loss, and the training process is stable by design, and insensitive to its initialization. Furthermore, the consistency of the approach is theoretically proven, and its flexibility and generalization efficiency are experimentally demonstrated: the same learned model can accurately handle unstructured meshes of various sizes, as well as different boundary conditions. To the best of our knowledge, $\Psi$-GNN is the first physics-informed GNN-based method that can handle various unstructured domains, boundary conditions and initial solutions while also providing convergence guarantees.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (70)
  1. Graph element networks: adaptive, structured computation and memory, in: International Conference on Machine Learning, PMLR. pp. 212–222.
  2. Layer normalization. arXiv preprint arXiv:1607.06450 .
  3. Deep equilibrium models. Advances in Neural Information Processing Systems 32.
  4. Stabilizing equilibrium models by jacobian regularization. arXiv preprint arXiv:2106.14342 .
  5. Interaction networks for learning about objects, relations and physics. Advances in neural information processing systems 29.
  6. Artificial neural networks approach for solving stokes problem. Applied Mathematics 1, 288.
  7. A multigrid tutorial. SIAM.
  8. A class of methods for solving nonlinear simultaneous equations. Mathematics of computation 19, 577–593.
  9. Physics-informed neural networks (pinns) for fluid mechanics: A review. Acta Mechanica Sinica , 1–12.
  10. A compositional object-based approach to learning physical dynamics. arXiv preprint arXiv:1612.00341 .
  11. Openfoam for computational fluid dynamics. Notices of the AMS 61, 354–363.
  12. A machine learning based solver for pressure poisson equations. Theoretical and Applied Mechanics Letters 12, 100362.
  13. Using neural networks to solve the 2d poisson equation for electric field computation in plasma fluid simulations. arXiv preprint arXiv:2109.13076 .
  14. Neural-network-based approximations for solving partial differential equations. communications in Numerical Methods in Engineering 10, 195–201.
  15. Deep statistical solvers. Advances in Neural Information Processing Systems 33, 7910–7921.
  16. Fast graph representation learning with pytorch geometric. arXiv preprint arXiv:1903.02428 .
  17. Physics-informed graph neural galerkin networks: A unified framework for solving pde-governed forward and inverse problems. Computer Methods in Applied Mechanics and Engineering 390, 114502.
  18. Gmsh: A 3-d finite element mesh generator with built-in pre-and post-processing facilities. International journal for numerical methods in engineering 79, 1309–1331.
  19. Understanding the difficulty of training deep feedforward neural networks, in: Proceedings of the thirteenth international conference on artificial intelligence and statistics, JMLR Workshop and Conference Proceedings. pp. 249–256.
  20. Eigenvalue computation in the 20th century. Journal of Computational and Applied Mathematics 123, 35–65.
  21. Recent advances in convolutional neural networks. Pattern recognition 77, 354–377.
  22. On stability and convergence of projection methods based on pressure poisson equation. International Journal for Numerical Methods in Fluids 26, 1039–1053.
  23. Convolutional neural networks for steady flow approximation, in: Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, pp. 481–490.
  24. Inductive representation learning on large graphs. Advances in neural information processing systems 30.
  25. Physics-embedded neural networks: E (n)-equivariant graph neural pde solvers. arXiv preprint arXiv:2205.11912 .
  26. Learning neural pde solvers with convergence guarantees. arXiv preprint arXiv:1906.01200 .
  27. A stochastic estimator of the trace of the influence matrix for laplacian smoothing splines. Communications in Statistics-Simulation and Computation 19, 433–450.
  28. Performance and accuracy assessments of an incompressible fluid solver coupled with a deep convolutional neural network. arXiv preprint arXiv:2109.09363 .
  29. Nsfnets (navier-stokes flow nets): Physics-informed neural networks for the incompressible navier-stokes equations. Journal of Computational Physics 426, 109951.
  30. Poisson surface reconstruction, in: Proceedings of the fourth Eurographics symposium on Geometry processing.
  31. Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907 .
  32. Machine learning–accelerated computational fluid dynamics. Proceedings of the National Academy of Sciences 118, e2101784118.
  33. Multilayer perceptrons and radial basis function neural network methods for the solution of differential equations: a survey. Computers & Mathematics with Applications 62, 3796–3811.
  34. Artificial neural networks for solving ordinary and partial differential equations. IEEE transactions on neural networks 9, 987–1000.
  35. Introduction to Numerical Methods for Variational Problems. doi:10.1007/978-3-030-23788-2.
  36. The finite element method: theory, implementation, and applications. volume 10. Springer Science & Business Media.
  37. Convolutional networks for images, speech, and time series. The handbook of brain theory and neural networks 3361, 1995.
  38. Neural algorithm for solving differential equations. Journal of Computational Physics 91, 110–131.
  39. Self-attention graph pooling, in: International conference on machine learning, PMLR. pp. 3734–3743.
  40. Neural operator: Graph kernel network for partial differential equations. arXiv preprint arXiv:2003.03485 .
  41. Multipole graph neural operator for parametric partial differential equations. Advances in Neural Information Processing Systems 33, 6755–6766.
  42. Simulating continuum mechanics with multi-scale graph neural networks. arXiv preprint arXiv:2106.04900 .
  43. Solving the finite-difference non-linear poisson–boltzmann equation. Journal of computational chemistry 13, 1114–1118.
  44. Newtonian limit of conformal gravity and the lack of necessity of the second order poisson equation. General Relativity and Gravitation 26, 337–361.
  45. Physics-informed neural networks for high-speed flows. Computer Methods in Applied Mechanics and Engineering 360, 112789.
  46. Numerical Solution of Partial Differential Equations: An Introduction. Cambridge University Press. URL: https://books.google.fr/books?id=0uf_iQIBFKgC.
  47. Multi-level gnn preconditioner for solving large scale problems. arXiv preprint arXiv:2402.08296 .
  48. Ds-gps: A deep statistical graph poisson solver (for faster cfd simulations). arXiv preprint arXiv:2211.11763 .
  49. Cfdnet: A deep learning-based accelerator for fluid simulations, in: Proceedings of the 34th ACM international conference on supercomputing, pp. 1–12.
  50. Introduction to Partial Differential Equations. Undergraduate Texts in Mathematics, Springer International Publishing. URL: https://books.google.fr/books?id=aQ8JAgAAQBAJ.
  51. Poisson cnn: Convolutional neural networks for the solution of the poisson equation on a cartesian mesh. Data-Centric Engineering 2.
  52. Automatic differentiation in pytorch .
  53. Learning mesh-based simulation with graph networks. arXiv preprint arXiv:2010.03409 .
  54. Hidden physics models: Machine learning of nonlinear partial differential equations. Journal of Computational Physics 357, 125–141.
  55. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational physics 378, 686–707.
  56. Physics-informed deep learning for incompressible laminar flows. Theoretical and Applied Mechanics Letters 10, 207–212.
  57. Introduction to the finite element method. McGraw-Hill Education.
  58. U-net: Convolutional networks for biomedical image segmentation, in: International Conference on Medical image computing and computer-assisted intervention, Springer. pp. 234–241.
  59. Iterative methods for sparse linear systems. SIAM.
  60. Learning to simulate complex physics with graph networks, in: International Conference on Machine Learning, PMLR. pp. 8459–8468.
  61. Graph networks as learnable physics engines for inference and control, in: International Conference on Machine Learning, PMLR. pp. 4470–4479.
  62. Adjustment of an inverse matrix corresponding to a change in one element of a given matrix. The Annals of Mathematical Statistics 21, 124–127.
  63. Modelling the dynamics of nonlinear partial differential equations using neural networks. Journal of Computational and Applied Mathematics 170, 27–58.
  64. Study on a poisson’s equation solver based on deep learning technique, in: 2017 IEEE Electrical Design of Advanced Packaging and Systems Symposium (EDAPS), IEEE. pp. 1–3.
  65. Solver-in-the-loop: Learning from differentiable physics to interact with iterative pde-solvers. Advances in Neural Information Processing Systems 33, 6111–6122.
  66. Graph attention networks. arXiv preprint arXiv:1710.10903 .
  67. Latent space physics: Towards learning the temporal evolution of fluid flow, in: Computer graphics forum, Wiley Online Library. pp. 71–82.
  68. Physics-informed machine learning approach for augmenting turbulence models: A comprehensive framework. Physical Review Fluids 3, 074602.
  69. A comprehensive survey on graph neural networks. IEEE transactions on neural networks and learning systems 32, 4–24.
  70. A convolutional neural network approach to training predictors for airfoil performance, in: 18th AIAA/ISSMO multidisciplinary analysis and optimization conference, p. 3660.
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