Intelligent optimization of airfoil motion using immersed boundary lattice Boltzmann method and deep reinforcement learning

This study proposes an active flow control strategy for an airfoil by integrating the immersed boundary–lattice Boltzmann method (IB-LBM) with the deep reinforcement learning (DRL) algorithm of Proximal Policy Optimization (PPO). The flow field is simulated using LBM, while the immersed boundary method is employed to accurately capture the interaction between the fluid and the moving airfoil. A PPO agent is trained to optimize the airfoil's motion in real time, with a reward function defined based on aerodynamic performance metrics, such as lift and drag coefficients. Numerical experiments are conducted under both steady and sinusoidal inflow conditions to assess the effectiveness and adaptability of the proposed control strategy. The results show that the PPO-controlled airfoil achieves substantial improvements in aerodynamic efficiency compared with uncontrolled cases, and the learned policy demonstrates robust transferability across different flow regimes. Overall, this work underscores the potential of coupling advanced computational fluid dynamics with DRL to tackle complex flow control problems and provides new insight for the intelligent optimization of wind energy systems.