Decomposition-Based Integer Programming for Coordinated Train Rerouting and Rescheduling

EasyChair Preprint 2603

Abstract

This paper deals with the coordinated train rerouting and rescheduling problem faced by railway infrastructure managers. Typically, the railway network is divided into non-overlapping control areas. Each control area is managed by a dispatcher, while control centers are in charge of coordinating the overall traffic. The rerouting and rescheduling problem arises when a perturbation occurs, i.e., when an unexpected event degrades operation. When this happens, the timetable cannot be operated as planned. Either timing or routes of trains have to be modified in order to minimize the impact of the perturbation and optimize a given objective, e.g., minimizing the total delay. This problem is known in the literature as real-time Railway Traffic Management Problem (rtRTMP). Several approaches have been proposed to deal with it. Differently, only few papers focus on the coordination of traffic management across control areas. In this paper, we aim to propose a general coordination framework in which the single dispatcher problem can be solved through virtually any approach among those proposed in the literature. Only the addition of some constraints or the slight modification of the objective function is considered possible to achieve coordination. This generality has the advantage of allowing the use of the most suitable approach for each control area, for example considering a microscopic or mesoscopic representation of the infrastructure depending on the layouts.

Keyphrases: Benders decomposition, Dispatching, coordination, mathematical programming

@booklet{EasyChair:2603,
  author    = {Peng Guo and Paola Pellegrini and Joaquin Rodriguez and Raffaele Pesenti},
  title     = {Decomposition-Based Integer Programming for Coordinated Train Rerouting and Rescheduling},
  howpublished = {EasyChair Preprint 2603},
  year      = {EasyChair, 2020}}