Learning physical models that can respect conservation laws

DOI10.1016/j.physd.2023.133952arXiv2302.11002MaRDI QIDQ6198203

Derek J. Hansen, Michael W. Mahoney, Danielle C. Maddix, Shima Alizadeh, Gaurav Gupta

Publication date: 21 February 2024

Published in: Physica D (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2302.11002

zbMATH Keywords

conservation lawspartial differential equationsuncertainty quantificationscientific machine learningphysically constrained machine learningshock location detection

Mathematics Subject Classification ID

Computational learning theory (68Q32) Bayesian inference (62F15) Bayesian problems; characterization of Bayes procedures (62C10) Sampling theory, sample surveys (62D05) Shocks and singularities for hyperbolic equations (35L67) Hyperbolic conservation laws (35L65) Finite difference methods for initial value and initial-boundary value problems involving PDEs (65M06) Finite volume methods for initial value and initial-boundary value problems involving PDEs (65M08) Finite volume methods for boundary value problems involving PDEs (65N08) PDE constrained optimization (numerical aspects) (49M41)

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Numerical artifacts in the discontinuous generalized porous medium equation: how to avoid spurious temporal oscillations
• Numerical simulation of dendritic solidification with convection: Two-dimensional geometry
• Fronts propagating with curvature-dependent speed: Algorithms based on Hamilton-Jacobi formulations
• Crystal growth and dendritic solidification
• A simple level set method for solving Stefan problems
• Numerical artifacts in the generalized porous medium equation: why harmonic averaging itself is not to blame
• When and why PINNs fail to train: a neural tangent kernel perspective
• Physics-informed neural networks for high-speed flows
• Conservative physics-informed neural networks on discrete domains for conservation laws: applications to forward and inverse problems
• Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
• The mimetic finite difference method for elliptic and parabolic problems with a staggered discretization of diffusion coefficient
• Finite Volume Methods for Hyperbolic Problems
• Neural tangent kernel: convergence and generalization in neural networks (invited paper)

This page was built for publication: Learning physical models that can respect conservation laws