Research on UAV Flight Trajectory Based on PSO and Improved PSO Algorithms

With the widespread application of UAV technology in military, agricultural, logistics, and other fields, path planning, as a core technology of autonomous navigation, faces challenges from complex environmental constraints and multi-objective optimization. This paper systematically investigates the application of Particle Swarm Optimization (PSO) and its improved strategies in UAV 3D path planning. Secondly, by comparing five algorithmstraditional PSO, Variable social weight PSO, mutated particle PSO, hybrid strategy PSO, and neural network-assisted PSOin complex terrains with varying numbers of peaks (N=5, 13, 20), their performance is evaluated. Simulation results indicate that the traditional PSO algorithm is simple and efficient but prone to premature convergence. The Variable social weight PSO excels in balancing exploration and exploitation capabilities, enhancing convergence speed. The mutated particle PSO and hybrid PSO effectively avoid local optima and demonstrate superior path quality in complex terrains. The neural network-assisted PSO incurs higher computational costs in high-dimensional complex environments and is susceptible to overfitting in simple environments. This study provides a theoretical basis for algorithm selection in different mission scenarios and proposes future directions for intelligent path planning technology development.