Ant colony optimization with immigrants schemes for the dynamic railway junction rescheduling problem with multiple delays

Eaton, J., Yang, S. and Mavrovouniotis, M. ORCID: 0000-0002-5281-4175, 2016. Ant colony optimization with immigrants schemes for the dynamic railway junction rescheduling problem with multiple delays. Soft Computing, 20 (8), pp. 2951-2966. ISSN 1432-7643

[img]
Preview
Text
6727_Mavrovouniotis.pdf - Published version

Download (1MB) | Preview

Abstract

Train rescheduling after a perturbation is a challenging task and is an important concern of the railway industry as delayed trains can lead to large fines, disgruntled customers and loss of revenue. Sometimes not just one delay but several unrelated delays can occur in a short space of time which makes the problem even more challenging. In addition, the problem is a dynamic one that changes over time for, as trains are waiting to be rescheduled at the junction, more timetabled trains will be arriving, which will change the nature of the problem. The aim of this research is to investigate the application of several different ant colony optimization (ACO) algorithms to the problem of a dynamic train delay scenario with multiple delays. The algorithms not only resequence the trains at the junction but also resequence the trains at the stations, which is considered to be a first step towards expanding the problem to consider a larger area of the railway network. The results show that, in this dynamic rescheduling problem, ACO algorithms with a memory cope with dynamic changes better than an ACO algorithm that uses only pheromone evaporation to remove redundant pheromone trails. In addition, it has been shown that if the ant solutions in memory become irreparably infeasible it is possible to replace them with elite immigrants, based on the best-so-far ant, and still obtain a good performance.

Item Type: Journal article
Publication Title: Soft Computing
Creators: Eaton, J., Yang, S. and Mavrovouniotis, M.
Publisher: Springer
Date: 2016
Volume: 20
Number: 8
ISSN: 1432-7643
Identifiers:
NumberType
10.1007/s00500-015-1924-xDOI
1924Publisher Item Identifier
Divisions: Schools > School of Science and Technology
Depositing User: Jonathan Gallacher
Date Added: 01 Dec 2016 11:32
Last Modified: 09 Jun 2017 14:08
URI: http://irep.ntu.ac.uk/id/eprint/29209

Actions (login required)

Edit View Edit View

Views

Views per month over past year

Downloads

Downloads per month over past year