Deep asymptotic expansion method for solving singularly perturbed time-dependent reaction-advection-diffusion equations

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhu, Qiao, Chaikovskii, Dmitrii, Jin, Bangti, Zhang, Ye
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918135941038080
author Zhu, Qiao
Chaikovskii, Dmitrii
Jin, Bangti
Zhang, Ye
author_facet Zhu, Qiao
Chaikovskii, Dmitrii
Jin, Bangti
Zhang, Ye
contents Physics-informed neural network (PINN) has shown great potential in solving partial differential equations. However, it faces challenges when dealing with problems involving steep gradients. The solutions to singularly perturbed time-dependent reaction-advection-diffusion equations exhibit internal moving transition layers with sharp gradients, and thus the standard PINN becomes ineffective. In this work, we propose a deep asymptotic expansion (DAE) method, which is inspired by asymptotic analysis and leverages deep learning to approximate the smooth part of the expansion. We first derive the governing equations for transition layers, which are then solved using PINN. Numerical experiments show that the DAE outperforms the standard PINN, gPINN, and PINN with adaptive sampling. We also show its robustness with respect to training point distributions, network architectures, and random seeds.
format Preprint
id arxiv_https___arxiv_org_abs_2505_23002
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Deep asymptotic expansion method for solving singularly perturbed time-dependent reaction-advection-diffusion equations
Zhu, Qiao
Chaikovskii, Dmitrii
Jin, Bangti
Zhang, Ye
Numerical Analysis
35B25, 65D17, 65N99, 68T07
Physics-informed neural network (PINN) has shown great potential in solving partial differential equations. However, it faces challenges when dealing with problems involving steep gradients. The solutions to singularly perturbed time-dependent reaction-advection-diffusion equations exhibit internal moving transition layers with sharp gradients, and thus the standard PINN becomes ineffective. In this work, we propose a deep asymptotic expansion (DAE) method, which is inspired by asymptotic analysis and leverages deep learning to approximate the smooth part of the expansion. We first derive the governing equations for transition layers, which are then solved using PINN. Numerical experiments show that the DAE outperforms the standard PINN, gPINN, and PINN with adaptive sampling. We also show its robustness with respect to training point distributions, network architectures, and random seeds.
title Deep asymptotic expansion method for solving singularly perturbed time-dependent reaction-advection-diffusion equations
topic Numerical Analysis
35B25, 65D17, 65N99, 68T07
url https://arxiv.org/abs/2505.23002