UAV logistics path planning based on hybrid PSO_GOOSE

Abstract

Unmanned Aerial Vehicles are gaining importance in logistics, since they can be easily and promptly deployed. Many logistics companies are now actively testing drone-based delivery solutions, hoping to make drones a regular tool in their delivery systems. In the near future, the logistics drone sector is expected to grow very quickly. However, path planning remains a critical challenge in Unmanned Aerial Vehicles logistics applications, as it must account for obstacle avoidance, path smoothness, and real-time responsiveness in complex environments. To achieve better global optimization performance and convergence reliability, a new hybrid algorithm PSO_GOOSE is proposed by combining Particle Swarm Optimization with the Goose algorithm in this study. Moreover, the Cauchy mutation is adopted to enhance the exploration ability, and the Greedy strategy is applied to speed up convergence.

Flight trajectory is represented and modelled as a spline-interpolated trajectory controlled by discrete waypoints in a two-dimensional grid environment. Through extensive simulation, the performance of the new algorithm is evaluated and compared with Particle Swarm Optimization, Ant Colony Optimization and A* algorithms. Five evaluation metrics are used: Best Cost, Path Length, Constraint Violation, Smoothness, and Runtime. The experiments show that the novel hybrid algorithm outperforms the baseline methods in key metrics, offering superior path feasibility, quality, and efficiency. Additionally, an ablation study confirms the individual contributions of Cauchy mutation and the greedy strategy to the overall algorithmic performance. The convergence behaviour and path visualizations further validate the robustness and adaptability of the proposed approach for logistics applications.