Physics-Informed Neural Networks for coupled stiff transport systems

DOI10.48550/arXiv.2606.21236MaRDI QIDQ7269968arXivFDO

Authors Laetitia Laguzet, Gabriel Turinici

Publication date 19 June 2026

Published in arXiv

Tldr This paper investigates the shortcomings of Physics-Informed Neural Networks on stiff hyperbolic systems like the Marshak wave equations and proposes targeted modifications (bounded activation, scale-aware loss, conservation law enforcement) leading to a novel framework that recovers reference solutions for these challenging problems.

Full work available at URL https://doi.org/10.48550/arXiv.2606.21236

arXiv classification

math.NAmath.OCphysics.comp-phphysics.flu-dyn

This page was built for publication: Physics-Informed Neural Networks for coupled stiff transport systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7269968)