Train scheduling is a crucial part of railway operations, where trains are allocated to particular routes and times to ensure a sustainable utilization of the railway network. Decisions of train scheduling are usually made in well advanced after considering various factors such as expected passenger and freight demand, infrastructure availability, and operational constraints. Nonetheless, scheduling extra train services, being it for passengers or freight, to accommodate unexpected demand in the railway system remains a persistent challenge. This is due to the fact that the existing timetable is typically designed with predetermined robustness, and the introduction of additional trains can adversely impact the robustness of the timetable. To address this issue, this paper presents an approach using an actor-critic reinforcement learning algorithm that takes into account the aspect of timetable robustness in solving the problem of scheduling additional trains. A case study of this method demonstrates that the proposed algorithm can take into account the importance of robustness in the decision-making process, resulting in well-informed decisions and a more reliable timetable.
Schedule Extra Train(s) into Existing Timetable Using Actor-Critic Reinforcement Learning
24.09.2023
666521 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Europäisches Patentamt | 2020
|Actor-Critic Reinforcement Learning for Control With Stability Guarantee
BASE | 2020
|