Bi-directional Control Characteristics of General Motors (GM) and Optimal Velocity Car- Following Models: Implications for Coordinated Driving in Connected Vehicle Environment

摘要

In natural traffic flow, the information from preceding vehicles determines driver behavior predominantly.With the availability of connected vehicle technologies (CVT), drivers can receive information from bothpreceding and following vehicles. This creates new opportunities for vehicle coordination and control atthe microscopic level based on bi-directional information. Although the bi-directional car followingmodels have been studied since the 1960s, most existing car-following models, especially those used byadaptive cruise control (ACC) technologies, are still forward-only car-following models. This paperserves as a first step towards the use of bi-directional car-following models for microscopic vehiclecoordination and control. The focus is on studying their general control characteristics and their impact ontraffic flow stability. A general bi-directional control framework is proposed to convert any car-followingmodel into bi-directional forms. Four representative GM (General Motors) and optimal velocity car-following models are reformulated and calibrated against field vehicle trajectory data collected in theNGSIM (Next Generation SIMulation) project. The bi-directional control characteristics of the selectedmodels are evaluated by tuning the percentage of the consideration of backward information in the finalcar-following decision. The evaluation uses forward versus backward acceleration diagrams and a ringroad stability analysis with respect to equilibrium states obtained from the NGSIM data. The increase ofbackward information contribution may help alleviate traffic congestion and stabilize traffic flow.Meanwhile, an operating range of backward information contribution between 5-20% is recommended sothat the resulting model can produce reasonable results for both free flow and congestion situation.