通信距离受限下无人机集群轨迹分布式滚动规划

For the security and efficiency issues of trajectory planning of unmanned aerial vehicle (UAV) swarms subject to the limited communication distance, this paper establishes the distributed receding planning framework and proposes the local-priority-decoupled sequential convex programming method (LPD-SCP). LPD-SCP divides swarm trajectory planning problems into a series of short-horizon single-UAV convex optimization subproblems, which can improve the computational efficiency and security of trajectories. This paper derives the length of feasible planning horizon to ensure the safety of trajectories. The communication-distance-limited local-priority decoupling mechanism is designed to adjust the priority of UAVs in the neighborhood according to the communication condition for achieving collision avoidance under communication distance limitation. In addition, the trajectory-time-consistency adjustment strategy is customized to realize the flight time consistency of UAVs in the neighborhood by updating the lower boundary constraint of flight time. It is theoretically guaranteed that the proposed LPD-SCP can obtain a swarm trajectory in several receding planning horizons subject to flight constraints. Simulation results demonstrate that LPD-SCP can plan the cooperative trajectories satisfying the flight performance, collision avoidance, and time consistency constraints under communication distance limitation, and the runtime of generating the short-horizon rendezvous trajectories for 15 UAVs is less than 4 s.