KTH-FlowAI/DRL_MARL_RayleighBenard3D_Control

Deep Reinforcement Learning (DRL) for the control of three-dimensional (3D) Rayleigh-Bénard convection (RBC) using the multi-agent reinforcement learning (MARL) framework

Control of three-dimensional Rayleigh-Bénard convection using multi-agent reinforcement learning

This is the official repository for Deep Reinforcement Learning (DRL) for the control of three-dimensional (3D) Rayleigh-Bénard convection (RBC) using the multi-agent reinforcement learning (MARL) framework. The DRL library used is Tensorforce and the numerical solver used for the CFD is the package shenfun

All code material for this repository will be released upon publication of our recently submitted manuscript.