01697654

Component

Urban rail transit operations are frequently subject to uncertain disruption events that may lead to large-scale deviations from planned travel patterns. In this paper, a two-stage MILP is formulated to optimize the planning strategies, using timetable planning and path choice planning as applications herein, that are robust against the worst-case platform disruption. In the planning stage, binary timetable decisions and path-based passenger flows on preferred paths are captured before the realization of the uncertain disruptions, which are modeled in a discrete uncertainty set. In the disruption stage, the disruption impact on passenger flows is evaluated by a linear arc-and-node based passenger routing model that simulates timetable-based passenger movement including queuing, boarding, alighting and transfering in a rail transit system, with the objective of minimizing aggregated system load in the aftermath of disruptions. The uncertain two-stage MILP is solved by a cutting plane algorithm based on Benders Decompositions. Computational experiments show that the robust timetable planning and path choice planning that are derived from the proposed framework significantly mitigate the worst deviations from original plans.

This paper was sponsored by TRB committee AP065 Standing Committee on Rail Transit Systems.

19-02405

English

8p

2019

Transportation Research Board 98th Annual Meeting

Digital/other

Figures; References

Linear programming; Planning; Rail transit; Service disruption

Operations and Traffic Management; Planning and Forecasting; Railroads

Transportation Research Board Annual Meeting 2019 Paper #19-02405

TRIS, TRB, ATRI

Dec 7 2018 9:33AM