Collision-Free Model Predictive Control for Periodic Trajectory Tracking of UAVs

This article focuses on the periodic trajectory tracking problem of UAVs in an environment with both static obstacles and moving obstacles, in which the reference trajectory allows to be unreachable. This problem is solved by integrating the trajectory planner with trajectory tracking controller in a single MPC optimization problem. So as to achieve the trajectory tracking task safely, additional variables are introduced into the MPC optimization problem to generate an optimal reachable trajectory to track. To be able to avoid collisions with obstacles, MINVO basis is used to obtain an outer polyhedral approximation of obstacles’ trajectories, and collision-free constraints are then derived by optimizing planes separating the polyhedrons from the trajectory of UAV. The MPC optimization formulated can be solved by a standard nonlinear programming solver. Finally, the efficacy of the proposed control algorithm is demonstrated by a simulation example for UAVs in a limited environment with obstacles.