Operations Research Models for Railway Rolling Stock Planning

Abstract

Chapter 2 describes the railway planning process at NS in detail. We also give a brief overview of operations research publications that address railway optimisation. Chapter 3 deals with tactical rolling stock circulations. We provide two models for this problem. First we describe the `Composition Model'. With appropriate ne-tuning, it can be solved to near optimality even for the hardest instances of NS within a couple of hours. This part of Chapter 3 is based on the paper Fioole, P.J., Kroon, L.G., Mar oti, G., Schrijver, A. (2004) A Rolling Stock Circulation Model for Combining and Splitting of Passenger Trains. CWI Research Report PNA{E0420, Center for Mathematics and Computer Science, Amsterdam, The Netherlands. To appear in European Journal of Operational Research. Also in Chapter 3, we describe the `Job Model' which is an alternative model for determining tactical rolling stock circulations. We compare the two models on instances of NS. In Chapter 4 we give two models for the maintenance routing problem that arises in short-term planning: the `Interchange Model' and the `Transition Model'. The Interchange Model is designed to take as much details of reality into account as possible. Besides complexity investigations, we propose a heuristic solution approach and report our computational results on instances of NS. This rst part of Chapter 4 is based on the paper 6 Chapter 1. Introduction Mar oti, G., Kroon, L.G. (2004). Maintenance Routing for Train Units: the Scenario Model. CWI Research Report PNA{E0414, Center for Mathematics and Computer Science, Amsterdam, The Netherlands. Revised version with the title Maintenance Routing for Train Units: the Interchange Model to appear in Computers and Operations Research. The Interchange Model requires a large amount of input data which may be inaccessible in real-life applications. This motivates the conceptually much simpler Transition Model. We discuss the computational complexity of the Transition Model and compare it with the Interchange Model based on computational results. This second part of Chapter 4 follows the paper Mar oti, G., Kroon, L.G. (2005) Maintenance Routing for Train Units: the Transition Model. Transportation Science, 39(4):518{ 525. Chapter 5 is devoted to operational rolling stock circulations. We describe a twophase approach that is used currently at NS for operational rolling stock planning. In this thesis we only study the second phase. We analyse the complexity of this problem and propose a heuristic solution method. Operational planning problems can also be formulated as instances of tactical rolling stock planning models. In Chapter 5, we compare the heuristic algorithm with the Composition Model on instances of NS. In Chapter 6 we draw some conclusions and indicate possible directions for further research on the topics considered in this thesis. This thesis is supplied with three appendices. Throughout the thesis we need terms for describing the railway planning and execution process. We provide our understanding of railway terminology in Appendix A. The notion of b-transshipments is used in Chapter 5; Appendix B gives the most important de nitions and theorems related to b-transshipments. Finally, the models in Chapters 3, 4 and 5 require a number of parameters and variables; all these notations are collected in Appendix C. %\ 2011-03-01 09:11:00 %\ 2014-03-13 10:43:00