Keywords: metro operations, timetable recovery, mixed-integer programming, skip-stop strategy, holding control, departure rescheduling, disturbance management, ridership weighting, headway regularity.

This study proposes an integrated mixed integer quadratic programming-based real-time timetable recovery model for high-frequency metro operations. The objective of our model is to minimise deviations between actual and planned headways, accounting for station importance based on ridership data. Upon disturbance, trains are classified as disturbed, already departed and not yet departed. Based on the classification, recovery measures such as skip-stop, holding and departure offset are applied. We applied this model to the Yokohama Municipal Subway Blue Line in Yokohama, Japan. In the experiments, we tested the model's performance by varying the disturbance location along the route and the number of rescheduled trains. It is observed that skip stops being a dominant recovery strategy effectively reducing the local deviations and improving the headway regularity by 30-49% depending upon the disturbance location, while holding and departure offsets can further improve the headway regularity to 54.6% in line with greater operational flexibility, compared to about 31.8% near terminal stations where recovery options are limited and based on the sensitivity analysis these improvement can be achieved by rescheduling 2-3 trains. The formulation is compared with metaheuristic algorithms, in which the model reliably finds globally optimal solutions in significantly shorter time.

download the full-text of this paper (PDF, 14 pages, 1112 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description