The bicycle-following model presented in this paper represents the behavior of bicyclists as they follow each other in traffic stream without an opportunity or sufficient stimulus to overtake or pass a lead bicycle. This model may be a significant component of a microscopic traffic simulation that represents transportation networks under a wide range of operating conditions. As traffic flow rate increases, the opportunity of vehicles to travel freely diminishes. Therefore, interaction between vehicles, including spacing and headway, speed differentials and the impact of acceleration and deceleration, play an increasingly significant role in determining the nature of traffic flow. Both bicyclist and her vehicle performance play a significant role in this interaction. The equations of motion may be used to represent the motion of unconstrained vehicles. However, for constrained vehicles, alternative descriptions of the interaction, including vehicle following and overtaking are needed. This paper presents a bicycle-following model tht was developed based on generalized estimating equations and bicycle trajectory data collected every second. The quasi-likelihood estimation method is used to account for the correlation between measurements on the same subject bicyclists. The model may be used in a traffic simulation to represent bicycles in traffic. It addresses the lack of studies on developing an understanding of bicyclist behavior.
Abstract