<ul><li>This paper presents a Reinforcement Learning (RL)-based control framework for position and attitude control of Unmanned Aerial Systems (UAS) subjected to significant disturbances.</li><li>The proposed method enables the system to anticipate and counteract disturbances by learning the relationship between the trigger signal and disturbance force.</li><li>Three policies were trained and evaluated: a baseline policy without exposure to disturbances, a reactive policy trained with disturbances but without the trigger signal, and a predictive policy that incorporates the trigger signal as an observation and is exposed to disturbances during training.</li><li>Simulation results demonstrate that the predictive policy outperforms the other policies by proactively correcting position deviations and improving UAS performance.</li></ul>

RL-based Control of UAS Subject to Significant Disturbance

Discover more