Integration, Decentralization and Self-Organization: Towards Better Public Transport
Abstract
Public transport brings undisputed benefits to modern-day societies. Aside from providing an affordable means to get around, its supreme efficiency in comparison with private transport plays a crucial role in curbing congestion and pollution, highlighting the importance of adequate design and operation of public transport systems.
The first part of this thesis seeks to improve the planning process of public transport operators by integrating planning steps that are traditionally performed sequentially. The first study considers a combination of line planning and vehicle scheduling, and presents methods that estimate how many vehicles are required to operate a line plan, without having to compute a timetable. The second study combines timetabling and vehicle scheduling, and develops a novel optimization model for jointly optimizing a periodic timetable and vehicle circulation schedule.
The second part of this thesis investigates decentralized strategies for operating public transport, with a focus on railway systems. Such strategies could be preferable over conventional centralized and schedule-based control in various scenarios. The first study in this part presents a theoretical analysis of a simple, decentralized strategy for dispatching vehicles. The second study considers the application of decentralized control to out-of-control situations in railways, which includes the development of a solution algorithm to find line plans that are suited for these circumstances. The final study in this thesis tests decentralized dispatching of both vehicles and crew in a microscopic railway simulation