Dynamic MPC Obstacle Avoidance Based Tube Sparse A* Path Planning Method for UAV Swarm Under Communication Delays

Cooperative path planning plays a vital role for the UAVs in real cooperative combat. However, the reliability and timeliness of the existing path planning methods are usually intractable to guarantee due to the existence of communication delays between UAVs. To deal with this issue, in this paper, considering the communication delays, a distinctive “global planning-local obstacle avoidance” cooperative path planning framework based on the integration of the tube sparse A* algorithm and model predictive control (MPC) is constructed. In global planning phase, the priority decoupling mechanism and receding planning framework are used to incorporate collision avoidance coordination as constraints for the path planning model. Considering the probabilistic distribution of adjacent UAVs’ positions caused by random communication delays, a real-time global tube sparse A* path planning algorithm is customized, in which communication delays are assumed to obey Markov distribution. Subsequently, in local collision avoidance phase, considering the uncertainty of communication delays, some path segments generated by global path planning may still be at risk of collision with adjacent UAVs, the path of UAVs inside the tube will be tracked and adjusted by solving the optimization problem of the distributed MPC model. Lastly, simulation results demonstrate the reliability and timeliness of proposed algorithms.