Scientific multi-agent reinforcement learning for wall-models of turbulent flows

The predictive capabilities of turbulent flow simulations, critical for aerodynamic design and weather prediction, hinge on the choice of turbulence models. The abundance of data from experiments and simulations and the advent of machine learning have provided a boost to turbulence modeling efforts....

Full description

Saved in:
Bibliographic Details
Published in:Nature communications 2022-03, Vol.13 (1), p.1443-1443, Article 1443
Main Authors: Bae, H. Jane, Koumoutsakos, Petros
Format: Article
Language:English
Subjects:
639/166
639/705
639/705/1042
Computer applications
Extreme values
Flow simulation
Fluid dynamics
Fluid flow
Humanities and Social Sciences
Large eddy simulation
Machine learning
Multiagent systems
multidisciplinary
Reinforcement
Reynolds number
Science
Science (multidisciplinary)
Simulation
Turbulence models
Turbulent flow
Vortices
Weather forecasting
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The predictive capabilities of turbulent flow simulations, critical for aerodynamic design and weather prediction, hinge on the choice of turbulence models. The abundance of data from experiments and simulations and the advent of machine learning have provided a boost to turbulence modeling efforts. However, simulations of turbulent flows remain hindered by the inability of heuristics and supervised learning to model the near-wall dynamics. We address this challenge by introducing scientific multi-agent reinforcement learning (SciMARL) for the discovery of wall models for large-eddy simulations (LES). In SciMARL, discretization points act also as cooperating agents that learn to supply the LES closure model. The agents self-learn using limited data and generalize to extreme Reynolds numbers and previously unseen geometries. The present simulations reduce by several orders of magnitude the computational cost over fully-resolved simulations while reproducing key flow quantities. We believe that SciMARL creates unprecedented capabilities for the simulation of turbulent flows. Simulations of turbulent flows are relevant for aerodynamic and weather modeling, however challenging to capture flow dynamics in the near wall region. To solve this problem, the authors propose a multi-agent reinforcement learning approach to discover wall models for large-eddy simulations.
ISSN:2041-1723
2041-1723
DOI:10.1038/s41467-022-28957-7