A New Design Method for Train Marshaling Evaluating the Transfer Distance of Locomotive

Chapter
Part of the Lecture Notes in Electrical Engineering book series (LNEE, volume 110)

Abstract

In this paper a new reinforcement learning system for generating marshaling plan of freight cars in a train is designed. In the proposed method, the total transfer distance of a locomotive is minimized to obtain the desired layout of freight cars for an outbound train. The order of movements of freight cars, the position for each removed car, the layout of cars in a train and the number of cars to be moved are simultaneously optimized to achieve the minimization of the transfer distance of locomotive. Initially, freight cars are located in a freight yard by the random layout, and they are removed and lined into a main track in a certain desired order in order to assemble an outbound train. Q-Learning is applied to reflect the transfer distance of the locomotive that is used to achieve one of the desired layouts in the main track. After adequate autonomous learning, the optimum schedule can be obtained by selecting a series of movements of freight cars that has the best evaluation.

Keywords

Scheduling Freight train Marshaling Q-Learning Container transfer problem 

References

  1. 1.
    Blasum, U., Bussieck, M.R., Hochstättler, W., Moll, C., Scheel, H.-H., Winter, T. (2000). Scheduling trams in the morning. Math. Meth. Oper. Res. 49(1):137–148.Google Scholar
  2. 2.
    Dahlhaus, E., Manne, F., Miller, M., Ryan, J. (2000). Algorithms for combinatorial problems related to train marshalling. In: Proceedings of the 11th Australasian workshop on combinatorial algorithms, pp 7–16.Google Scholar
  3. 3.
    Eggermont, C., Hurkens, C.A.J., Modelski, M., Woeginger, G.J. (2009). The hardness of train rearrangements. Oper. Res. Lett. 37:80–82.CrossRefMATHMathSciNetGoogle Scholar
  4. 4.
    He, S., Song, R., Chaudhry, S.S. (2000). Fuzzy dispatching model and genetic algorithms for railyards operations. Eur. J. Oper. Res. 124(2):307–331.CrossRefMATHGoogle Scholar
  5. 5.
    Hirashima, Y. (2011). A new rearrangement plan for freight cars in a train: Q-learning for minimizing the movement counts of freight cars. Lecture notes in electrical engineering, 70 LNEE, pp 107–118.Google Scholar
  6. 6.
    Hirashima, Y. (2011). A new reinforcement learning method for train marshaling based on the transfer distance of locomotive. Lecture notes in engineering and computer science: proceedings of the international multiconference of engineers and computer scientists 2011, IMECS 2011, Hong Kong, March 2011, 16–18, pp 80–85.Google Scholar
  7. 7.
    Jacob, R., Marton, P., Maue, J., Nunkesser, M. (2007). Multistage methods for freight train classification. In: Proceedings of 7th workshop on algorithmic approaches for transportation modeling, optimization, and systems, pp 158–174.Google Scholar
  8. 8.
    Kroon, L.G., Lentink, R.M., Schrijver, A. (2008). Shunting of passenger train units: an integrated approach. Transport. Sci. 42:436–449.CrossRefGoogle Scholar
  9. 9.
    Sutton, R.S., Barto, A.G. (1999). Reinforcement learning. MIT Press, MA.Google Scholar
  10. 10.
    Tomii, N., Jian, Z.L. (2000). Depot shunting scheduling with combining genetic algorithm and pert. In: Proceedings of 7th international conference on computer aided design, manufacture and operation in the railway and other advanced mass transit systems, pp 437–446.Google Scholar
  11. 11.
    Watkins, C.J.C.H., Dayan, P. (1992). Q-learning. Mach. Learn. 8:279–292.MATHGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2012

Authors and Affiliations

  1. 1.Osaka Institute of TechnologyHirakataJapan

Personalised recommendations