Robust timetables for railway traffic: Development and evaluation of an optimization-based method

Important results from the project

The project aims to improve reliability and punctuality of public railway transportation by developing and evaluating an optimization-based method for creating robust timetables. Robust timetables refer here to timetables which contain time margins on strategic places which enable the trains to follow their schedule even if certain delays occur. Thereby are other trains affected less and critical connections and rolling-stock and staff schedules can be maintained as much as possible.

Expected long term effects

The project intends to enable a reduced sensitivity to disturbances in the railway network and thereby also enabling a more sustainable use of the network capacity. The results are intended to serves as guidance to railway operators on how to insert cost-effective buffer times in their timetables.

Approach and implementation

The project runs from 2010-04-01 until 2011-12-31. The project group consists of Ph. D. Johanna Törnquist Krasemann, Ph.D. Anders Peterson, and a new Ph D student at Linköping University. In addition, personnel from BV and SJ will participate. Cooperation with the KTH- projectet Simulering för tidtabellsplanering för att minska förseningarna will also take place. The project is divided into ten work packages (WP): WP1 serve to plan the project in more detail and define resources from BV and SJ which need to be allocated. A reference group will also be formed. In WP2 the part of the network and the traffic which will be part of the study is defined along with input data required. The hypotheses in focus will also be further described. In WP3, an analysis of existing related approaches and research in the area of buffer time insertion for railway traffic timetabling is carried out. In WP4, we calibrate and adjust our model and compare it to the ones applied in the KTH-project. The model will also be evaluated by comparison to real scenarios in discussion with relevant personnel at BV and SJ. In WP5 the method is applied in simulated experiments and based on the results it is evaluated in WP6. Based on conclusions from WP5 and 6, the method is further developed, applied and evaluated in WP7-WP9. In WP10 the project is finalised and documentation prepared. The results from the project are, however, disseminated continuously.