In progress Improving Reliability For Railway Passengers

Reference:
ERIM PhD 2016 ESE DH AW

Abstract

In this project we focus on the rescheduling of rolling stock and the
rescheduling of the associated shunting plans, in particular for the case
of smaller disruptions on the day of operation. First of all, we will adapt
existing solution methodologies for rolling stock rescheduling for this real-time setting. In this respect, we will focus on rolling stock rescheduling for delays, a type of disruption that is often encountered in this setting. Moreover, we investigate the use of heuristics for solving the rescheduling problem for small disruptions. Secondly, we focus on the shunting problem faced by dispatchers in the rescheduling phase. Instead of focusing on the parking of train units at the shunting yard, we focus on a more integral problem that combines nding a location to park train units with findings crews to execute the chosen shunting plans and finding a path through the infrastructure to allow the shunting movement. To do so, we will first investigate solution methodologies for these problems and will the investigate how these can be integrated into the existing rolling stock rescheduling problems.

Keywords

Rolling Stock Rescheduling, Shunting, Disruption Management, Public Transport Optimization

Time frame

2016 - 2020

Topic

In disruption management, the timetable, rolling stock and crew schedules are modified. Algorithms to support these tasks have been developed for some of these processes. However, steps like shunting of rolling stock and dealing with the limited infrastructure within the stations is usually ignored in these algorithms. Therefore, we would like to develop new models and algorithms for these aspects within this PhD project.

Moreover, passenger flows and passenger behaviour are hardly considered in current models and algorithms. In this PhD project, we would like to extend current rescheduling models by considering objectives as minimizing passenger delays. As a consequence, the underlying algorithms have to be modified as well.

The PhD project is part of a larger series of 6 project co-financed by Netherlands Railways (Nederlandse Spoorwegen, or shortly NS). Cooperation with other PhD students and researchers will be stimulated. In addition, the student is expected to work at least one day a week in the head office of NS such that he/she can obtain enough practical knowledge about his/her topic.

Supervisory Team

Dennis Huisman
Professor of Public Transport Optimization
  • Promotor