Abstract

In this article, we consider the problem of planning preventive maintenance of railway signals in Denmark. This case is particularly relevant as the entire railway signalling system is currently being upgraded to the new European Railway Traffic Management System (ERTMS) standard. This upgrade has significant implications for signal maintenance scheduling in the system. We formulate the problem as a multi-depot vehicle routing and scheduling problem with time windows and synchronisation constraints, in a multi-day time schedule. The requirement that some tasks require the simultaneous presence of more than one engineer means that task synchronisation must be considered. A multi-stage constructive framework is proposed, which first distributes maintenance tasks using a clustering formulation. Following this, a Constraint Programming (CP) based approach is used to generate feasible monthly plans for large instances of practical interest. Experimental results indicate that the proposed framework can generate feasible solutions and schedule a monthly plan of up to 1000 tasks for eight crew members, in a reasonable amount of computational time.

Highlights

  • The European Rail Traffic Management System (ERTMS) (Bloomfield, 2006) is the new generation of rail communication and control signalling systems introduced by the European Union

  • The contribution of this article is twofold: 1. We show that the Preventive Signalling Maintenance Crew Scheduling Planning (PSMCSP) can be formulated as a Multi Depot Vehicle Routing and Scheduling Problem (MDVRSP) with synchronisation constraints

  • We report the results of experiments using the stage-based solution approach described in Sections 3 and 4 for set of test cases covering a number of scenarios

Read more

Summary

Introduction

The European Rail Traffic Management System (ERTMS) (Bloomfield, 2006) is the new generation of rail communication and control signalling systems introduced by the European Union. This article focuses on a crew scheduling and routing problem at the tactical level, arising in the planning of preventive maintenance tasks to be performed on signals geographically spread across the rail network. To tackle the interdependency problem in the presence of exact synchronisation constraints several approaches have been suggested These include allowing infeasibility in partial solutions during the search (Oertel, 2000; De Rosa et al, 2002; Wen, Larsen, Clausen, Cordeau, & Laporte, 2009; Prescott-Gagnon, Desaulniers, & Rousseau, 2014), intensification of the search space indirectly in local search and large neighbourhood search (Lim, Rodrigues, & Song, 2004; Li, Lim, & Rodrigues, 2005), and approximation of the cost function (De Rosa et al, 2002; Wen et al, 2009). We show that the Preventive Signalling Maintenance Crew Scheduling Planning (PSMCSP) can be formulated as a Multi Depot Vehicle Routing and Scheduling Problem (MDVRSP) with synchronisation constraints.

Maintenance planning in ERTMS
Requirement for clustering the maintenance region
MIP formulation
Proposed solution framework
First stage
Second stage
Third stage
Fourth stage
Routing and scheduling phase
Route interdependency
The problem as a CSP
Adding constraints
Solving the routing and scheduling problem
Feasibility check
Updating and merging abstract day IDs
Experimental results
Test case description
Comparison with a commercial MIP solver
Main results
Individual cluster results
Optimality gap
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call