A Disturbance Rejection Control Method for Carrier Landing Based on DDPG Algorithm

Landing control for carrier-based unmanned aerial vehicles (UAV) is highly challenging due to severe environment disturbance. This paper proposes a reinforcement learning-based anti-disturbance approach to solve these problems. A relative motion model is established, integrating UAV and carrier dynamics, air wake disturbances, and deck motion, along with an Automatic Carrier Landing System (ACLS). A robust trajectory tracking controller is designed using the Deep Deterministic Policy Gradient (DDPG) algorithm to improve the landing system's precision and reliability. The proposed control framework comprises three components: the UAV actuation system, flight dynamics model, and the DDPG-based controller. Numerical simulations were conducted to evaluate performance under oscillating carrier deck motion, and the landing accuracy of the proposed method is compared with that of the other methods. A comparative analysis has been conducted, and the results indicate that the DDPG-based approach exhibits enhanced robustness and accuracy, effectively mitigating disturbances and ensuring precise landings. This study validates the proposed strategy as a promising solution for UAV carrier landing in challenging conditions.