An exact and heuristic framework for rolling stock rescheduling with railway infrastructure availability constraints

Disruptions on the railway network can lead to reduced availability of the railway infrastructure, which requires rolling stock dispatchers to adjust the planning of the rolling stock. In this paper, we develop fast rolling stock rescheduling methods which ensure feasibility with respect to the availability of the railway infrastructure. In particular, we explore the option of performing shunting movements at stations where shunting does not take place in current practice, due to the large number of trains that pass through or due to the complexity of the station layout. We introduce an exact rolling stock rescheduling algorithm and an iterative heuristic, which alternate between two mathematical formulations, namely one that creates an interim rolling stock schedule and one that tries to fit the suggested shunting movements between the remaining railway traffic. We test our solution approach with instances that contain complete railway blockages on the Dutch railway network. We successfully identify feasible shunting movements and find an average improvement in the objective function of 19% over the rolling stock schedule that would be obtained if performing shunting movements at the considered stations is prohibited. • We consider rolling stock rescheduling with infrastructure availability constraints. • We develop an exact method which resembles a Combinatorial Benders’ approach. • We develop an iterative heuristic which alternates between two MIP formulations. • Computational experiments based on real-life instances from Netherlands Railways.