Comparing the effectiveness of two real-time train rescheduling systems in case of perturbed traffic conditions

摘要

Rescheduling train traffic in a busy and complex railway area is a challenging task, partly because of the high number of constraints to be taken into account, and partly because of the many variables involved. Currently this task is performed almost exclusively by human traffic operators. Previous attempts to provide an automated decision support system have been limited to identifying and solving train conflicts locally. Recently innovative dispatching support tools have been presented that are able to cope with large (real-time) timetable perturbations, such as train delays and their propagation. However, there is a lack of computational studies that underline their additional practical value.rnIn this paper we compare two advanced support systems for real-time rescheduling of train operations, developed for the German and Dutch railway networks. The research aim is to establish a bench mark for future co-operation and exchange of innovative solutions. A common test case from the Dutch railway network, the dispatching area between Utrecht and Den Bosch, and disturbed traffic conditions are studied to evaluate the two dispatching support tools in terms of delay minimization. Since these tools make use of different mathematical optimization techniques for the computation of running times and train sequences, a detailed comparison of the proposed rescheduling solutions is provided. The use of railway capacity is illustrated in order to enable an easy and fast detection of the conflicts between the trains running in the network and to get precise information about their resolution by the different rescheduling techniques.