Evaluation of a car-following model using systems dynamics

Models that describe the processes by which drivers follow each other in the traffic stream are generally referred to as car-following models. During the past 50 years, driver behavior within the traffic stream has been studied and models that attempt to describe this behavior have been proposed. Car-following models have frequently been developed for the purpose of incorporation with microscopic simulation models. These models have then been used to evaluate a wide range of potential geometric options, operational strategies, and/or policies. In this paper we formulate a car-following model using the systems dynamic (SD) approach. We compare the behavior of the proposed SD model to the GM model, a classic car-following model that has been extensively described in the literature. These comparisons illustrate that the proposed SD car-following model avoids several unrealistic characteristics of the GM model. The ultimate objective is to use the proposed model to investigate the mechanisms leading to rear-end crashes and to quantify the impact that different technologies or policies (e.g. driver vision enhancement systems) may have on rear-end crash potential.