Optimization of Multi-UAV Trajectories in Complex Obstacle Scenarios

In the problem of multi-objective point task allocation and trajectory optimization for multi-UAV clusters in complex obstacle environments, an improved genetic algorithm (INSGA-II) is employed to solve the task sequences of the multi-UAVs. The genetic process considers the effect of the average congestion of the parent individuals, and obstacle effects are incorporated into the mission planning constraints, with the total distance of the UAVs from the obstacles as an optimization objective. A multi-segment Bezier trajectory optimization method is proposed to quickly obtain a flight trajectory that conforms to the dynamics of a fixed-wing UAV after determining the UAV cluster task sequence, ensuring continuity between the multi-segment trajectories. Simulations conducted in a two-dimensional scenario with obstacles show that the INSGA-II algorithm converges to a better sequence more quickly than the random search algorithm, and the multi-segment Bezier trajectory optimization method efficiently generates trajectories that meet the dynamic constraints according to the task sequence.