Saved in:
Bibliographic Details
Main Authors: Hao, Baoli, Braga-Neto, Ulisses, Liu, Chun, Wang, Lifan, Zhong, Ming
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2404.16189
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911196230189056
author Hao, Baoli
Braga-Neto, Ulisses
Liu, Chun
Wang, Lifan
Zhong, Ming
author_facet Hao, Baoli
Braga-Neto, Ulisses
Liu, Chun
Wang, Lifan
Zhong, Ming
contents Training Physics-Informed Neural Networks (PINNs) on stiff time-dependent PDEs remains highly unstable. Through rigorous ablation studies, we identify a surprisingly critical factor: the enforcement of initial conditions. We present the first systematic ablation of two core strategies, hard initial-condition constraints and adaptive loss weighting. Across challenging benchmarks (sharp transitions, higher-order derivatives, coupled systems, and high frequency modes), we find that exact enforcement of initial conditions (ICs) is not optional but essential. Our study demonstrates that stability and efficiency in PINN training fundamentally depend on ICs, paving the way toward more reliable PINN solvers in stiff regimes.
format Preprint
id arxiv_https___arxiv_org_abs_2404_16189
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Stability in Training PINNs for Stiff PDEs: Why Initial Conditions Matter
Hao, Baoli
Braga-Neto, Ulisses
Liu, Chun
Wang, Lifan
Zhong, Ming
Numerical Analysis
Training Physics-Informed Neural Networks (PINNs) on stiff time-dependent PDEs remains highly unstable. Through rigorous ablation studies, we identify a surprisingly critical factor: the enforcement of initial conditions. We present the first systematic ablation of two core strategies, hard initial-condition constraints and adaptive loss weighting. Across challenging benchmarks (sharp transitions, higher-order derivatives, coupled systems, and high frequency modes), we find that exact enforcement of initial conditions (ICs) is not optional but essential. Our study demonstrates that stability and efficiency in PINN training fundamentally depend on ICs, paving the way toward more reliable PINN solvers in stiff regimes.
title Stability in Training PINNs for Stiff PDEs: Why Initial Conditions Matter
topic Numerical Analysis
url https://arxiv.org/abs/2404.16189