Trajectory Tracking with Obstacle Avoidance for Autonomous UAV Swarms Based on Distributed Model Predictive Control

Multi-agent quadcopter systems, a specialized class of unmanned aerial vehicles (UAVs), have become key players in various industries, showing potential for further growth. However, the inherently nonlinear and complex behavior of quadcopters, coupled with the need for collision and obstacle avoidance within the swarm, while simultaneously achieving collective goals, poses significant challenges for efficient control. Trajectory tracking - a particularly demanding task for swarms - requires a delicate balance between precise tracking and safe navigation. In this work, we address the challenge of trajectory tracking for multi-agent quadcopter configurations through a novel distributed model predictive control (DMPC) framework, with individual local controllers for each agent. For a balanced control scheme, we employ both PID controllers for angle regulation and linear adaptive MPC (LAMPC) controllers for three-dimensional position control. The local MPC schemes ensure safe trajectory tracking, without the need to use predetermined reference trajectories for each agent and predefined formation strategies. Simulation results of the proposed framework demonstrate advanced robustness and dynamic adaptation to unpredicted situations.