A Hybrid GA-APF 3D Trajectory Planning Method for Large-Scale UAV Formation Systems

In the field of trajectory planning, the generation of a feasible trajectory ensures flight safety for unmanned aerial vehicle (UAV) formations and minimizes mission execution costs, making it a crucial and actively researched topic. To generate a feasible trajectory in large-scale UAV formation systems, this paper proposes a novel Hybrid Genetic Algorithm-Artificial Potential Field (GA-APF) method for trajectory planning. The genetic algorithm (GA) is modified by introducing a new selector and mutation operator to enhance the efficiency of assigning mission targets. Furthermore, the artificial potential field (APF) method is improved by incorporating a three-layered repulsive potential field and modifying the calculation of the attractive potential field to ensure safe and accessible paths for all UAVs. Finally, the trajectory is smoothed through the use of the B-Spline algorithm. Simulation experiment on the Hybrid GA-APF method using 1000 UAVs shows that the method not only reduces the trajectory length by more than 60% but also generates safer trajectories. Using trajectories generated by the method on a real-world flight platform can enable large-scale UAV formations to complete their missions safely.