Car-driver models for manual and automated travel

摘要

This paper presents a model for manual and automated traffic flow. The model is based on the abstraction of vehicle activities by the space and time taken up by the vehicle engaged in the activity. The manual driver model includes the activities of vehicle following and lane change and allows for different levels of aggressiveness of drivers. The parameters of the activites are calibrated with real highway data. The expectation is that this model will provide more realistic estimates of highway capacity while retaining the efficiency of macro-scale simulators.