The stochastic multiple depot electric vehicle scheduling problem with recourse

Abstract

In this paper, a heuristic is presented to improve the robustness of a given railway timetable in a bottleneck area. The timetable can be adapted by adjusting the timing and routing of trains, but cannot deviate too much from the current timetable. Robustness is measured with an objective function that considers the buffer times between train pairs. The developed algorithm updates the routing and the timing in separate steps. It is applied to a bottleneck area of the Belgian network. The results show that the objective can be improved by about 10% when alternative routing options are considered. Additional experiments with smaller instances indicate that the heuristic is capable of finding near-optimal solutions.