Lingyun Meng

A Review of Online Dynamic Models and Algorithms for Railway Traffic Management

Railway timetables are developed to make operations robust and resilient to small delays. However, disturbances perturb the daily plan, and dispatchers adjust the plan to keep operations feasible and to limit delay propagation. Rescheduling approaches aim at updating the offline timetable at best, in the presence of delays. We present a survey of the recent approaches on online railway traffic rescheduling problems, which exhibit dynamic and stochastic (or, at least, not completely deterministic) aspects. In fact, while online static rescheduling has reached a wide degree of dissemination, much is still to be done with regard to the probabilistic nature of the railway traffic rescheduling problems, and also how to best take uncertainty into account for future states. Open challenges for the future research are finally outlined.

A review of online dynamic models and algorithms for railway traffic control

We present a survey of the recent approaches on railway transportation problems, that exhibit dynamic and stochastic (or, at least, not completely deterministic) aspects. In fact, while deterministic and static (or single-shot) systems have reached a wide degree of dissemination, much is still to be done with regard to the online nature of the railway problems; and how to best take into account for uncertainty. We review approaches for four dynamic interrelated problems such as train path rescheduling (called also dispatching, rescheduling, disposition), crew rescheduling, rolling stock rescheduling, and the issue of determining updated passenger connection plans (delay management). Open challenges for the research are outlined.

Integrated stochastic optimization approaches for tactical scheduling of trains and railway infrastructure maintenance

Abstract This work addresses a tactical railway traffic scheduling problem focused on the optimization of train sequencing and routing decisions and timing decisions related to short-term maintenance works in a railway network subject to disturbed process times. This is modeled as a mixed-integer linear programming formulation in which the traffic flow and track maintenance variables, constraints and objectives are integrated under a stochastic environment. The resulting bi-objective optimization problem is to minimize the deviation from a scheduled plan and to maximize the number of aggregated maintenance works under stochastic disturbances. The two objectives require to schedule competitive train operations versus maintenance works on the same infrastructure elements. Computational experiments are performed on a realistic railway network. We measure the quality of the integrated solutions in terms of their robustness to stochastic perturbations of the train travel times and of the maintenance works. Pareto optimal methods are compared for the bi-objective problem. We also evaluate the impact of introducing routing stability constraints in order to force the trains to keep the same route among the different stochastic disturbed scenarios. The experiments show that forcing the routing stability reduces the routing flexibility and the ability to optimize the two performance indicators when dealing with stochastic disturbances.