A new idea for train scheduling using ant colony optimization

This paper develops a new algorithm to the train scheduling problem using Ant Colony System (ACS) meta-heuristic. At first, a mathematical model for a kind of train scheduling problem is developed and then the algorithm based on the meta-heuristic is presented to solve the problem. The problem is considered as a traveling salesman problem (TSP) wherein cities represent the trains. ACS determines the sequence of trains dispatched on the graph of the TSP. Based on the sequence obtained and removing for collisions incurred, train scheduling is determined. Numerical examples in small and medium sizes are solved by using the algorithm. The solutions are compared to the exact optimum solutions to check for quality and accuracy. Comparison of the solutions shows that the algorithm results in good enough quality and time savings. A case study is presented to illustrate the solution.