Êtes-vous un étudiant de l'EPFL à la recherche d'un projet de semestre?
Travaillez avec nous sur des projets en science des données et en visualisation, et déployez votre projet sous forme d'application sur Graph Search.
Bus operations throughout the world are increasingly being equipped with Intelligent Transport Systems such as Automatic Vehicle Location (AVL). AVL can support a variety of functions, including Dynamic Bus Fleet Management (DBFM), which has yet to be established in most bus fleets in the UK in a systematic way. To develop a fundamental understanding of the bus and traffic related incidents in bus-based public transport for DBFM purposes, a microscopic simulation model, capable of modelling these incidents, has been developed and applied to a variety of scenarios. This paper describes the design and development of the model Simulating Incidents for BUs FlEet Management (SIBUFEM), for modelling bus operations during full-day periods in which incidents of different types can occur. The paper describes the model's functionality, including the use of journey time profiles, passenger-dependent bus stop dwell times and deterministic time-dependent queuing theory. Results focus on key performance measures including, but not limited to, bus journey times, passenger waiting times and bus delays resulting from various bus and traffic incidents. SIBUFEM has been applied to a main bus corridor in Southampton, UK, with bus flows increased to produce a high-frequency service. A base case of ‘normal’ operations has been established, for comparison with results from a number of incident scenarios, using key model performance parameters of average bus journey time, bus speed and excess waiting time. Incidents range from bus breakdowns, to traffic incidents such as road-works, traffic accidents and illegal parking. In SIBUFEM, these are specified in terms of their location, duration and severity (loss of capacity). The paper concludes with a discussion on potential DBFM strategies and how the SIBUFEM can be further developed to allow these strategies to be evaluated