Analyse expérimentale du comportement des conducteurs dans un carrefour urbain sans feux (Experimental analysis of driver behaviour at unsignalized urban intersections)

Unsignalized intersections are the most frequent type in an urban network and consequently have an important role. Drivers receive no instructions on such intersections and must decide when to enter the intersection, looking for a safe gap in which to enter the traffic stream. Time gaps are random and the gap acceptance process is the outcome of an interaction between gap demand and supply: gap demand consists of seeking sufficiently large time gaps to enter the intersection while gap supply is the way such intervals become available to the driver. This paper presents the findings of a study of driver behaviour at an unsignalized urban intersection, and these results should be seen as being more concerned with providing an initial understanding of the phenomena than developing a specific tool. The principal objectives of the study were to analyze the respective influences on the performance of manoeuvres of environmental factors and driver-related factors, and to investigate the potential of disaggregate models for simulating driver behaviours. After a stateof-the-art survey of traffic simulation models for intersections, the paper describes a field survey which was followed by a laboratory study of the behavioural parameters of drivers. Finally, the univariate and multivariate statistical analyses, conducted using the resulting database, are presented.