Mathematics

Deep reinforcement learning (DRL) is a mathematical framework that has been used to design and learn control policies in different domains, and several applications in physics research have been proposed, as well. Here we introduce a reinforcement learning (RL) environment to design control strategies in turbulent fluid flows enclosed in a channel.DRL leverages the high-dimensional data that can be sampled from flow simulations to achieve challenging objectives such as the reduction of the skin-friction coefficient or the enhancement of certain coherent structures features.Our results show that DRL control policies outperform classical opposition control in drag reduction and they can successfully promote highly elongated velocity streaks.These applications highlight the potential of DRL for both flow control and numerical experiment design for scientific discovery.