Prioritization of trains

-

Railroad infrastructure has a number of highly utilized sections through which many trains have to pass. Administrative rules enforced by the infrastructure manager (Jernbaneverket) help to sequence the train traffic. Unfortunately, these rules are static and cannot take into account consequences of prioritizing a certain train for other train traffic. Also, they do not work well for a congested system such that deviations from the rules are frequent to get all trains through a bottleneck without too much delay. However, the sequencing of trains at one location will also affect sequencing and utilization of other bottlenecks. Delays may quickly propagate throughout the day and the network. For a manual dispatcher it is very hard to assess the effects of a local decision on other parts of the network and other times of the day. This project will develop optimization based decision support tools for the dispatchers at the infrastructure manager, helping them to find good decisions for sequencing trains thr ough a portfolio of bottlenecks. The underlying mathematical model will be based on maximizing socio-economic utility rather than just focusing on improving the utilization of some bottlenecks. We extend previous work to look at more than one bottleneck at a time. This ensures that trains get priority either because this is more beneficial to society as a whole or because it improves punctuality and regularity and, thus, the value of a portfolio of some other trains. Another significant difference to pr evious work is that we combine economic theory with operations research methods. This will help to create better sequencing decisions but makes it also possible to later connect these decisions to incentive mechanisms aimed at the actors in this industry. The decision support tool is planned to be used as part of the daily operations in the Norwegian railway system, resulting in higher infrastructure utilization and improved on-time performance.