Integrated optimization of train scheduling from planning to operation

Background

As tactical plans, railway train timetables are programmed and updated every year or every season to define the network routes, the arrival and departure times, and the train orders at passing stations. Generally, the train timetabling problem has a goal of improving efficiency of rail networks in terms of higher infrastructure occupation and better service quality (e.g., a lower travel time).When a train timetable is put into practice, various sources of perturbations may influence the train movements, thus causing primary delays and even secondary delays to the planned train timetable. The (on-line) train rescheduling problem aims at updating the planned timetable at best by taking proper dispatching measures, which can adjust the impacted schedules from perturbations (i.e., make operations feasible) and further reduce potential negative consequences (e.g., limit delay propagation).

It has been widely recognized that the (off-line) planned timetables have significant impact on the quality of the (on-line) rescheduled timetables. The interaction between these two has been mostly studied with an aim of designing a robust timetable in the planning stage, which is able to handle minor disturbances and make operations resilient to small delays in operational stage. Such a joint consideration of off-line train scheduling and on-line train rescheduling can provide a global view from planning trains to operating trains, and explore the interaction between these two and the impact of off-line schedules on on-line operations. Moreover, the integrations of train (re-)scheduling and other aspects, e.g., preventive maintenance time slots planning, train control, signalling commands, and competitive equity, have practical significance, but mostly absent in the literature. This brings about the purpose of this research, i.e., (re-)scheduling trains from the planning to operational stage, by using the integrated optimization of jointly considering preventive maintenance time slots planning, train control, signalling impacts, and competitive equity.

Objective

The research focuses on the train (re-)scheduling problem from planning to operation, integrating the preventive maintenance time slots planning problem, train control problem, signalling commands, and competitive equity into the train (re-)scheduling problem by using optimization methods and distributed optimization methods.

Project team members

• X. Luan (Xiaojie)
• prof.dr.ir. B. De Schutter (Bart) 
• dr.ir. T.J.J. van den Boom (Ton)
• prof.dr. F. Corman (Francesco), Institute for Transport Planning and Systems (IVT), ETH Zurich
• prof.dr.ir. G. Lodewijks (Gabriel), School of Aviation, Faculty of Science, University of New South Wales
• prof.dr. L. Meng (Lingyun), School of traffic and transportation, Beijing Jiaotong University
• prof.dr. Jianrui Miao, State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University
• dr. Yihui Wang, State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University

Keywords

Train (re-)scheduling, Preventive maintenance planning, Train control, Signalling systems, Equity

Sponsored by

China Scholarship Council (CSC)