It is widely-recognized that driving aggressiveness affects traffic efficiency and emissions. The paper analyzes the quantitative impacts of different car-following behaviors on traffic efficiency, fuel consumptions, and emissions. Using real world vehicle trajectory data, Gipps’ car-following behaviors are first characterized by Genetic Algorithm based on microscopic traffic flow parameters. According to K-means clustering methods, driving behavior can be categorized into three types, ‘aggressive’, ‘normal’ and ‘timid’ driving behavior. Then simulation experiment is developed to analyze the fuel economy impacts of driving aggressiveness. The simulation set up 120 vehicles with three different driving types on a flat single-lane highway, consisting of dozens of scenarios. The proportion of aggressive driving behavior increases from 0% to 100% with a gap of 10%, the timid driving behavior also increased from 0% to the highest rate at a specific ratio of aggressive driving behavior. When the proportion of ‘aggressive’ and ‘timid’ driver behavior are determined, the proportion of normal drivers is also determined. The simulation result shows the fuel consumption would increase with the higher ratios of aggressive drivers.

CICTP