To improve time-efficient trajectory planning for robotic arms and address limitations like premature convergence in standard PSO, this paper presents a robot trajectory planning method based on Velocity Pausing Particle Swarm Optimization is proposed. The method integrates a dynamic velocity pause strategy and adaptive parameter adjustment to balance global search with local refinement. We apply this algorithm to the IRB-2600 robot, building a trajectory model based on hybrid cubic-quintic polynomial interpolation under kinematic constraints. Results indicate that VPPSO significantly enhances convergence speed and optimization accuracy. Compared with traditional planning and classical PSO, VPPSO shortens trajectory duration by 47.74% and 18.55%, respectively, without compromising feasibility.