CAR FOLLOWING MODELS - REVIEW AND FUTURE POTENTIALS

摘要

The concept of the car following theory, the manner in which two cars follow one another without passing, has been closely studied in the last decades. As applied to traffic engineering and safety research, many models were developed to best mimic the interaction between adjacent vehicles in a traffic stream. The car following theory also forms the process of the microscopic traffic simulation models which attempt to replicate driver behavior in a traffic stream. Though most existing applications are limited to historical traffic data, wide utilization of new data collection tools, such as Global Position System and Automatic Vehicle Identification technologies are expected to lead to rapid improvements in existing car following models. The large amount of real-time traffic data provides a possibility to better calibrate and validate car following models to better represent the traffic conditions. This study summarizes the commonly used car following models, and discusses the applications and limitations of each. Using current limitations, this study also discusses the future potential improvements of car following models with new technologies, such as the real-time traffic information provided by the intelligent transportation technologies.