With implementation of a newly developed 'Cruise-Assist Highway System', one major impact would be replacement of normal (no-assistance) vehicle with vehicle instrumented with assistance device. Secondly, the influence of behaviour of the vehicle instrumented on driving behaviours of normal vehicles surrounding it should also be examined. To evaluate these two kinds of impact, driving behaviour models for normal vehicles, used in traffic simulation, must be established. There are many existing car-following models, but they cannot describe real traffic such as capacity bottleneck phenomena occurred at sag sections and/or tunnel entrances. This study reveals the model structure that could describe such traffic phenomena, and the method to evaluate the parameters of the model.
Abstract