Abstract
Railway traffic is the biggest individual electricity consumer in Germany, amounting to 2% of the country’s total electricity usage. However, up to 20% of the annual electricity cost depends on the highest power value drawn within the billing period. In this paper, we optimize the timetables of railway traffic in order to avoid high peaks in power consumption, while preserving at the same time some usability and safety considerations. We propose an exact mixed integer programming model together with a systematic way of simplifying the model in order to obtain feasible solutions that are not far from the optimum. We also discuss two possible dynamic programming approaches that may be used for solving small instances with a specific structure. Our approach became Team Optimixtli’s winning entry in the Discrete Optimization Challenge: Energy-Efficient Train Timetables. This competition was part of the Open Research Challenge 2015 organized by the Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) in Germany.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
- 2.
A black dot in the second figure means that our solution exceeds this instantaneous maximum.
References
Albrecht, T.: Reducing power peaks and energy consumption in rail transit systems by simultaneous train running time control. Comput. Railw. IX pp. 885–894 (2010)
Friedrich-Alexander-Universität Erlangen-Nürnberg: Discrete Optimization: Energy-efficient Train Timetables (2015). https://openresearchchallenge.org/challengers2015
Gurobi Optimization, Inc.: Gurobi Optimizer Reference Manual (2016). http://www.gurobi.com
Haahr, J., Kidd, M.: Modeling and solution methods for the energy-efficient train timetables problem, Technical report. Technical University of Denmark (2015)
Hansen, I., Pachl, J.: Railway Timetabling and Operations: Analysis - Modelling - Optimisation - Simulation - Performance Evaluation. Eurail press (2014)
Ilgmann, G.: Die Bahn im Klima-Test. Frankfurter Allgemeine Zeitung (2007). http://www.faz.net/aktuell/wissen/energieverbrauch-die-bahn-im-klima-test-1494443.html
Sans, B., Girard, P.: Instantaneous power peak reduction and train scheduling desynchronization in subway systems. Transp. Sci. 31(4), 312–323 (1997)
Acknowledgements
We thank the Mexican National Council on Science and Technology (CONACyT) for providing scholarships to most students in our graduate program (including two team members) and also for providing a scholarship to another team member through the National System of Researchers (SNI). We thank the Mexiquense Council on Science and Technology (COMECyT) for providing a scholarship to our last team member. We thank Universidad Autónoma Metropolitana Azcapotzalco for funding Research Project SI004-13 (Algorithms and Models for Network Optimization Problems) and also for allowing us the use of some computing facilities through the Systems Department. We thank Gurobi Optimization for giving us free licenses to their software. We thank the Friedrich-Alexander-Universität Erlangen-Nürnberg for organizing the Discrete Optimization Challenge and the participating teams for their great effort. Last, but not least, we thank honorary team member Gabrijela Zaragoza for proposing a nice team name, a portmanteau of optimization and the nahuatl word mixtli (cloud).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Appendix
Appendix
In what follows, we display power consumption in two figures for each instance: the first for the given schedule, the second for our best result (Figs. 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24).
As before, a red dot means power consumption, a blue dot means power generation. However, to avoid clutter, each dot represents an average power consumption over one minute. The vertical scale accommodates the maximum instantaneous power consumption as given in the current schedule,Footnote 2 while the horizontal scale accommodates the 17 fifteen-minutes intervals. This scale holds for both figures.
We also display in light red the average power consumption (and in light blue the average unused power generation) on each fifteen-minutes interval. In this case, the vertical scale accommodates the maximum average power generation on a fifteen-minutes interval as given in the current schedule. This scale also holds for both figures.
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG
About this chapter
Cite this chapter
Castro Campos, R.A., Pérez Pérez, S.L., Vazquez Casas, G., Zaragoza Martínez, F.J. (2018). Mixed Integer Programming Formulation for the Energy-Efficient Train Timetables Problem. In: Maldonado, Y., Trujillo, L., Schütze, O., Riccardi, A., Vasile, M. (eds) NEO 2016. Studies in Computational Intelligence, vol 731. Springer, Cham. https://doi.org/10.1007/978-3-319-64063-1_3
Download citation
DOI: https://doi.org/10.1007/978-3-319-64063-1_3
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-64062-4
Online ISBN: 978-3-319-64063-1
eBook Packages: EngineeringEngineering (R0)