Simultaneous locomotive assignment and train scheduling on a single-track railway line: A simulation-based optimization approach

Abstract

• Two optimization models for integrated problem are formulated. • We propose an efficient approach to assign locomotives and schedule trains. • Numerical experiments validate the effectiveness of the solution approach. This paper aims to assign locomotives to unscheduled trains with pre-planned departure times instead of trains with a pre-scheduled timetable on a single-track railway line. Considering both rigorous and relaxed situations, this paper formulates two mixed-integer programming models of the integrated optimization models, to simultaneously assign locomotives to trains and schedule trains, so that train delay is minimized. In order to solve these problems within reasonable computational times, we develop an efficient optimization approach on the basis of designed locomotive assignment rules and train movement simulation methods. For comparison purposes, two heuristics, including local branching method and relaxation induced neighborhood search approach which can be embedded into an optimization solver to effectively solve mixed-integer programming models, are also introduced. Finally, we implement extensive numerical experiments to validate the dependability and effectiveness of the proposed optimization models and approaches.