Train re-scheduling with genetic algorithms and artificial neural networks for single-track railways

Dundar S., Şahin İ.

TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES, vol.27, pp.1-15, 2013 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 27
  • Publication Date: 2013
  • Doi Number: 10.1016/j.trc.2012.11.001
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.1-15
  • Keywords: Train re-scheduling, Train dispatching, Conflict resolution, Genetic algorithms, Binary encoding, Artificial neural networks, MODELS
  • Yıldız Technical University Affiliated: Yes


Train re-scheduling problems are popular among researchers who have interest in the railway planning and operations fields. Deviations from normal operation may cause inter-train conflicts which have to be detected and timely resolved. Except for very few applications, these tasks are usually performed by train dispatchers. Due to the complexity of re-scheduling problems, dispatchers utilize some simplifying rules to resolve conflicts and implement their decisions accordingly. From the system effectiveness and efficiency point of view, their decisions should be supported with appropriate tools because their immediate decisions may cause considerable train delays in future interferences. Such a decision support tool should be able to predict overall implications of the alternative solutions. Genetic algorithms (GAs) for conflict resolutions were developed and evaluated against the dispatchers' and the exact solutions. The comparison measures are the computation time and total (weighted) delay due to conflict resolutions. For benchmarking purposes, artificial neural networks (ANNs) were developed to mimic the decision behavior of train dispatchers so as to reproduce their conflict resolutions. The ANN was trained and tested with data extracted from conflict resolutions in actual train operations in Turkish State Railways. The GA developed was able to find the optimal solutions for small sized problems in short times, and to reduce total delay times by around half in comparison to the ANN (i.e., train dispatchers). (C) 2012 Elsevier Ltd. All rights reserved.