Developing advanced route choice models for heavy goods vehicles using GPS data

摘要

GPS data has established itself as a key tool for measuring route choices and the subsequent modellingof these choices in random utility models. This paper presents an application that is somewhatdierent in scope from past work, looking at the modelling of route choices for heavy goods vehicles,which typically make longer journeys and where the decision making is potentially underpinned bydierent priorities.Furthermore, while many previous route choice studies have been conducted atthe level of metropolitan areas or at best small countries, the present application uses the entire roadnetwork of England.The study showed that superior model performance was obtained by making useof the error components specication put forward by Frejinger and Bierlaire (2007) which capturescorrelation between routes sharing key roads. A set of illustrative forecasting runs also revealedlow elasticities in response to changes in travel time, reecting the limited opportunity for avoidingspecic roads on long distance journeys by heavy goods vehicles.