Railway operators adjust timetables, and accordingly reschedule rolling stock circulation and crew duties, when the train operations are disrupted by accidents or adverse weather conditions. This paper discusses the problem of rescheduling driver assignment to freight trains after timetable adjustment has been completed. We construct a network from the disrupted situation, and model the problem as an integer programming problem with set-covering constraints combined with set-partitioning constraints. The integer program is solved by column generation in which we reduce the column generation subproblem to a shortest path problem and such paths by utilizing data parallelism. Numerical experiments using a real timetable, driver scheduling plan and major disruption data in the highest-frequency freight train operation area in Japan reveal that our method provides a quality driver rescheduling solution within 25 seconds.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Keisuke SATO, Naoto FUKUMURA, "Real-Time Freight Train Driver Rescheduling during Disruption" in IEICE TRANSACTIONS on Fundamentals,
vol. E94-A, no. 6, pp. 1222-1229, June 2011, doi: 10.1587/transfun.E94.A.1222.
Abstract: Railway operators adjust timetables, and accordingly reschedule rolling stock circulation and crew duties, when the train operations are disrupted by accidents or adverse weather conditions. This paper discusses the problem of rescheduling driver assignment to freight trains after timetable adjustment has been completed. We construct a network from the disrupted situation, and model the problem as an integer programming problem with set-covering constraints combined with set-partitioning constraints. The integer program is solved by column generation in which we reduce the column generation subproblem to a shortest path problem and such paths by utilizing data parallelism. Numerical experiments using a real timetable, driver scheduling plan and major disruption data in the highest-frequency freight train operation area in Japan reveal that our method provides a quality driver rescheduling solution within 25 seconds.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/transfun.E94.A.1222/_p
Copy
@ARTICLE{e94-a_6_1222,
author={Keisuke SATO, Naoto FUKUMURA, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Real-Time Freight Train Driver Rescheduling during Disruption},
year={2011},
volume={E94-A},
number={6},
pages={1222-1229},
abstract={Railway operators adjust timetables, and accordingly reschedule rolling stock circulation and crew duties, when the train operations are disrupted by accidents or adverse weather conditions. This paper discusses the problem of rescheduling driver assignment to freight trains after timetable adjustment has been completed. We construct a network from the disrupted situation, and model the problem as an integer programming problem with set-covering constraints combined with set-partitioning constraints. The integer program is solved by column generation in which we reduce the column generation subproblem to a shortest path problem and such paths by utilizing data parallelism. Numerical experiments using a real timetable, driver scheduling plan and major disruption data in the highest-frequency freight train operation area in Japan reveal that our method provides a quality driver rescheduling solution within 25 seconds.},
keywords={},
doi={10.1587/transfun.E94.A.1222},
ISSN={1745-1337},
month={June},}
Copy
TY - JOUR
TI - Real-Time Freight Train Driver Rescheduling during Disruption
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 1222
EP - 1229
AU - Keisuke SATO
AU - Naoto FUKUMURA
PY - 2011
DO - 10.1587/transfun.E94.A.1222
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E94-A
IS - 6
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - June 2011
AB - Railway operators adjust timetables, and accordingly reschedule rolling stock circulation and crew duties, when the train operations are disrupted by accidents or adverse weather conditions. This paper discusses the problem of rescheduling driver assignment to freight trains after timetable adjustment has been completed. We construct a network from the disrupted situation, and model the problem as an integer programming problem with set-covering constraints combined with set-partitioning constraints. The integer program is solved by column generation in which we reduce the column generation subproblem to a shortest path problem and such paths by utilizing data parallelism. Numerical experiments using a real timetable, driver scheduling plan and major disruption data in the highest-frequency freight train operation area in Japan reveal that our method provides a quality driver rescheduling solution within 25 seconds.
ER -