Analysis of car-following behavior using microscopic trajectory data

摘要

The development of accurate and robust models in the field of car-following has suffered greatly from the lack of appropriate microscopic data. Due to this, little is known about differences in car-following behavior between individual drivers. This paper therefore studies car-following behavior of individuals using vehicle trajectory data that is extracted from images collected from a helicopter. The analysis was performed by estimating parameters of different specifications of the well-known GHR car-following rule for individual drivers. This analysis showed that in 95% of the cases, a statistical relation between the stimuli and the response could be established. The main contribution of this paper is however that considerable differences between the car-following behavior of individual drivers could be identified. These differences are expressed in terms of different optimal parameter values for the reaction time and the sensitivity, as well as different car-following models that appear to be optimal based on individual driver data.