Physics-Informed Graph-Mesh Networks for PDEs: A hybrid approach for complex problems

Abstract

The recent rise of deep learning has led to numerous applications, including solving partial differential equations using Physics-Informed Neural Networks. This approach has proven highly effective in several academic cases. However, their lack of physical invariances, coupled with other significant weaknesses, such as an inability to handle complex geometries or their lack of generalization capabilities, make them unable to compete with classical numerical solvers in industrial settings. In this work, a limitation regarding the use of automatic differentiation in the context of physics-informed learning is highlighted. A hybrid approach combining physics-informed graph neural networks with numerical kernels from finite elements is introduced. After studying the theoretical properties of our model, we apply it to complex geometries, in two and three dimensions. Our choices are supported by an ablation study, and we evaluate the generalisation capacity of the proposed approach.

References in corpus (13)

• Adam: A Method for Stochastic Optimization
• A Comprehensive Survey on Graph Neural Networks
• An Energy Approach to the Solution of Partial Differential Equations in Computational Mechanics via Machine Learning: Concepts, Implementation and Applications
• On the eigenvector bias of Fourier feature networks: From regression to solving multi-scale PDEs with physics-informed neural networks
• A unified deep artificial neural network approach to partial differential equations in complex geometries
• Hamiltonian Neural Networks
• Graph Neural Networks with convolutional ARMA filters
• Modeling the Dynamics of PDE Systems with Physics-Constrained Deep Auto-Regressive Networks
• Physics-informed graph neural Galerkin networks: A unified framework for solving PDE-governed forward and inverse problems
• GraphCast: Learning skillful medium-range global weather forecasting
• Deep Theory of Functional Connections: A New Method for Estimating the Solutions of PDEs
• Exact conservation laws for neural network integrators of dynamical systems
• Physics-Informed Graph Convolutional Networks: Towards a generalized framework for complex geometries