An Improved Joint Particle Swarm and Ant Colony Optimization Method for High-Speed Train Scheduling

An improved joint particle swarm-ant colony optimization (PS-ACO) method is proposed for the dynamic scheduling of high-speed trains with operation delays. Firstly, on the basis of the established dynamic train scheduling model, this paper utilizes the particle swarm algorithm to reduce the search scope quickly, and the search results can be used as the initial pheromone component of ant colony algorithm, which can avoid the problem of wide search range in the early stage of ant colony algorithm. Then, the ant colony algorithm is proposed to achieve the narrow window and refined search, which improves the accuracy of solution and obtains a better actual train operation diagram. The total train delay time is minimized and the dynamic train scheduling problem is effectively solved with PS-ACO method. Finally, with the simulation examples and comparative analysis, the effectiveness of the joint PS-ACO method is verified.