Learning robust marking policies for adaptive mesh refinement
Preprint
Abstract
In this work, we revisit the marking decisions made in the standard adaptive finite element method (AFEM). Experience shows that a na\"{i}ve marking policy leads to inefficient use of computational resources for adaptive mesh refinement (AMR). Consequently, using AFEM in practice often involves ad-hoc or time-consuming offline parameter tuning to set appropriate parameters for the marking subroutine. To address these practical concerns, we recast AMR as a Markov decision process in which refinement parameters can be selected on-the-fly at run time, without the need for pre-tuning by expert users. In this new paradigm, the refinement parameters are also chosen adaptively via a marking policy that can be optimized using methods from reinforcement learning. We use the Poisson equation to demonstrate our techniques on $h$- and $hp$-refinement benchmark problems, and our experiments suggest that superior marking policies remain undiscovered for many classical AFEM applications. Furthermore, an unexpected observation from this work is that marking policies trained on one family of PDEs are sometimes robust enough to perform well on problems far outside the training family. For illustration, we show that a simple $hp$-refinement policy trained on 2D domains with only a single re-entrant corner can be deployed on far more complicated 2D domains, and even 3D domains, without significant performance loss. For reproduction and broader adoption, we accompany this work with an open-source implementation of our methods.Keywords
All Related Versions
This publication has 9 references indexed in Scilit:
- Efficient Time-Stepping for Numerical Integration Using Reinforcement LearningSIAM Journal on Scientific Computing, 2023
- MFEM: A modular finite element methods libraryComputers & Mathematics with Applications, 2021
- A data-driven framework for error estimation and mesh-model optimization in system-level thermal-hydraulic simulationNuclear Engineering and Design, 2019
- Lower bounds of the discretization error for piecewise polynomialsMathematics of Computation, 2013
- Neural network time series forecasting of finite-element mesh adaptationNeurocomputing, 2005
- A Convergent Adaptive Algorithm for Poisson’s EquationSIAM Journal on Numerical Analysis, 1996
- Toward a universal adaptive finite element strategy part 3. design of meshesComputer Methods in Applied Mechanics and Engineering, 1989
- Toward a universal adaptive finite element strategy, part 2. A posteriori error estimationComputer Methods in Applied Mechanics and Engineering, 1989
- On an h-type mesh-refinement strategy based on minimization of interpolation errorsComputer Methods in Applied Mechanics and Engineering, 1985