Railway crew scheduling with grouping evolutionary algorithm

Railway crew scheduling furnishes time schedules for crews who provide services along the trip. In this paper, with emphasis on scheduling of head of crews for coach trains, a crew scheduling problem is presented in a network form with task arcs. To solve the problem, a meta heuristic algorithm based on grouping evolutionary algorithm is developed. The grouping evolutionary algorithm contains two search methods which generate offspring from a parent chromosome; one of these methods is based on a logic constraints heuristic algorithm, and the other one is relied on branch and bound approach. Computational results showed that combining grouping evolutionary algorithm and branch and bound approach generates good solutions, and for the problems that we knew their optimal solutions accomplish optimal outcomes. It is also shown that the algorithm provides good results for large scale real world problems.