As an optimization tool, PSO has the advantages of being simple and easy to achieve. The advantages of swarm intelligence and fast convergence make it suitable for large-scale complex network optimization problems. This paper analyses several strategies to improve particle swarm optimization, identify the advantages relative to the standard algorithm, and seizes the fast convergence characteristics of PSO. Through adjusting the algorithm parameters and the algorithm structure, it compensates the defect of the algorithm’s easy to fall into local optimum defect effectively, thereby increase its global search capability. Through proper selection of the correspondence of solution spaces and particles of PSO, it applies PSO to Open pit road network routing problem successfully.