Establishment of the Rayleigh Benard baseline flow used at the start of the Deep Reinforcement Learning episodes. Starting from the fluid at rest and a uniform temperature gradient between the bottom wall (T=2) and the top wall (T=1), the flow reaches a stable state where two Rayleigh Benard convection cells are observed.

Video prepared by Joel Vasanth, for more information, see our paper at: [ Ссылка ]

List of videos in the series:
- Part 1/4: baseline convergence: [ Ссылка ]
- Part 2/4: SARL controller: [ Ссылка ]
- Part 3/4: MARL controller: [ Ссылка ]
- Part 4/4: coalesced convection cell: [ Ссылка ]

свернуть