Model-based deep reinforcement learning for accelerated learning from flow simulations

Andre Weiner, Janis Geise

In recent years, deep reinforcement learning has emerged as a technique to solve closed-loop flow control problems. Employing simulation-based environments in reinforcement learnin…