Acoustics-based Active Control of Unsteady Flow Dynamics using Reinforcement Learning Driven Synthetic Jets

Abstract: This study proposes the use of deep reinforcement learning (DRL) to actively control wakes and noise from flow past a cylinder by leveraging acoustic-based pressure feedback. A hydrophone array captures downstream signals, enabling a DRL agent to adjust jet actuators on the cylinder's surface in real-time. The method reduces noise levels by up to 9.5\% and drag coefficients by up to 23.8\%, effectively minimizing flow-induced vibrations. This highlights the potential of DRL-driven active flow control for engineering applications.