Optimizing Quadcopter Trajectory Tracking with Deep-Q Reinforcement PID Controller in Uncertain Environments

Autonomous PID tuning is a compelling yet challenging application given the complicated dynamics of quadcopters, especially in unsafe environments. This study presents innovative PID variants, including Deep Q-Reinforcement Learning (RL-Q) PID, which aim to improve the stabilization of quadcopters. Using MATLAB, a comparative analysis with a conventional PID controller to determine their respective performance. To thoroughly evaluate their effectiveness, simulations are performed for two different trajectories-a circular and a dynamic spiral path. The simulation outcomes consistently reveal the superior performance of RL-Q PID compared to the standard PID. The RL-Q PID consistently outperforms, showcasing its advanced capability to stabilize the quadcopter on challenging traj ectories. These results provide valuable insights for the further development of control algorithms for quadcopter trajectory tracking and highlighting the performance of the proposed RL-Q PID, particularly in navigating uncertain and dynamic environments.