MeshDQN: A Deep Reinforcement Learning Framework for Improving Meshes in Computational Fluid Dynamics

TL;DR

MeshDQN introduces a versatile deep reinforcement learning framework that automatically coarsens CFD meshes, reducing manual effort and computational costs while maintaining simulation accuracy across different flow regimes.

Contribution

It presents a general-purpose graph neural network-based deep Q network for adaptive mesh coarsening in CFD, requiring only a single prior simulation and no assumptions about flow or mesh type.

Findings

Successfully improves meshes for two 2D airfoils.

Reduces need for extensive training data and multiple simulations.

Operates without assumptions on flow regime or solver.

Abstract

Meshing is a critical, but user-intensive process necessary for stable and accurate simulations in computational fluid dynamics (CFD). Mesh generation is often a bottleneck in CFD pipelines. Adaptive meshing techniques allow the mesh to be updated automatically to produce an accurate solution for the problem at hand. Existing classical techniques for adaptive meshing require either additional functionality out of solvers, many training simulations, or both. Current machine learning techniques often require substantial computational cost for training data generation, and are restricted in scope to the training data flow regime. MeshDQN is developed as a general purpose deep reinforcement learning framework to iteratively coarsen meshes while preserving target property calculation. A graph neural network based deep Q network is used to select mesh vertices for removal and solution interpolation is used to bypass expensive simulations at each step in the improvement process. MeshDQN requires a single simulation prior to mesh coarsening, while making no assumptions about flow regime, mesh type, or solver, only requiring the ability to modify meshes directly in a CFD pipeline. MeshDQN successfully improves meshes for two 2D airfoils.