Saved in:
Bibliographic Details
Main Authors: Huergo, David, Alonso, Laura, Joshi, Saumitra, Juanicoteca, Adrian, Rubio, Gonzalo, Ferrer, Esteban
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2407.15872
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929431843438592
author Huergo, David
Alonso, Laura
Joshi, Saumitra
Juanicoteca, Adrian
Rubio, Gonzalo
Ferrer, Esteban
author_facet Huergo, David
Alonso, Laura
Joshi, Saumitra
Juanicoteca, Adrian
Rubio, Gonzalo
Ferrer, Esteban
contents We explore a reinforcement learning strategy to automate and accelerate h/p-multigrid methods in high-order solvers. Multigrid methods are very efficient but require fine-tuning of numerical parameters, such as the number of smoothing sweeps per level and the correction fraction (i.e., proportion of the corrected solution that is transferred from a coarser grid to a finer grid). The objective of this paper is to use a proximal policy optimization algorithm to automatically tune the multigrid parameters and, by doing so, improve stability and efficiency of the h/p-multigrid strategy. Our findings reveal that the proposed reinforcement learning h/p-multigrid approach significantly accelerates and improves the robustness of steady-state simulations for one dimensional advection-diffusion and nonlinear Burgers' equations, when discretized using high-order h/p methods, on uniform and nonuniform grids.
format Preprint
id arxiv_https___arxiv_org_abs_2407_15872
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A reinforcement learning strategy to automate and accelerate h/p-multigrid solvers
Huergo, David
Alonso, Laura
Joshi, Saumitra
Juanicoteca, Adrian
Rubio, Gonzalo
Ferrer, Esteban
Machine Learning
Computational Physics
We explore a reinforcement learning strategy to automate and accelerate h/p-multigrid methods in high-order solvers. Multigrid methods are very efficient but require fine-tuning of numerical parameters, such as the number of smoothing sweeps per level and the correction fraction (i.e., proportion of the corrected solution that is transferred from a coarser grid to a finer grid). The objective of this paper is to use a proximal policy optimization algorithm to automatically tune the multigrid parameters and, by doing so, improve stability and efficiency of the h/p-multigrid strategy. Our findings reveal that the proposed reinforcement learning h/p-multigrid approach significantly accelerates and improves the robustness of steady-state simulations for one dimensional advection-diffusion and nonlinear Burgers' equations, when discretized using high-order h/p methods, on uniform and nonuniform grids.
title A reinforcement learning strategy to automate and accelerate h/p-multigrid solvers
topic Machine Learning
Computational Physics
url https://arxiv.org/abs/2407.15872