Abstract: Aiming at the problems of traditional particle swarm optimization algorithms easily converging to local optima and low search efficiency, an improved algorithm is proposed and applied to the path planning problem of Automated Guided Vehicle (AGV). Firstly, non-linear decreasing inertia weights are introduced to adjust the influence of particles themselves on optimization at different stages. Then, adaptive improvements are made to the two learning factors to enhance the algorithm's local and global search capabilities. Finally, a fitness function considering path length and smoothness is proposed, and local optimal regions are eliminated by interfering with particle velocity to improve the quality of the search path. Multiple experiments were conducted in four environments with different map sizes and obstacle complexity. Simulation results showed that the improved algorithm reduced the average search path by 7.9% and the average iteration times by 20.2% compared to the original algorithm, demonstrating superior path planning capabilities.

Key words: path planning, particle swarm optimization algorithm, inertial weight, learning factors