Study on mixed traffic of autonomous vehicles and human-driven vehicles with different cyber interaction approaches

Abstract

The emergence of autonomous vehicles will significantly improve traffic efficiency and safety. Before the fully autonomous driving of traffic system, the mixed traffic with autonomous and human-driven vehicles will be a typical feature for the future road traffic. To study the properties of a mixed traffic mode combining autonomous and human-driven vehicles, a cellular automata model based traffic system considering autonomous and human-driven vehicles is proposed. In the mixed traffic model, human-driven vehicles can use multiple preceding vehicles' information within drivers' visual range for vehicle motion. For autonomous vehicles, the interaction mechanism of information is iterative and the communication range for vehicle-to-vehicle is considered. The results from this study show that the average traffic flow can be enhanced with the increase of communication range for autonomous vehicles. But when the communication range is big enough, its role on traffic flow is little. Also, it is found that human-driven vehicles' behavior of obtaining multiple preceding vehicles' information can significantly impact the meta-stable state of traffic flow in the average flow density phase. Further, the relationship between the average flow and vehicle density of the mixed traffic flow with random distribution, uniform distribution and separate distribution is discussed. And it reveals that the traffic capacity of the mixed traffic flow with separate distribution is more impressive comparing to the traffic capacity with random distribution and uniform distribution. Finally, the traffic congestion rate under different values of penetration rate is verified and it uncovers that the congestion rate is reduced with the increase of penetration rate.