Graph Representation of Road and Traffic for Autonomous Driving

Jianglin Qiao, Dongmo Zhang, Dave de Jonge

Abstract: Autonomous driving has the potential to radically change the way vehicles interact each other. This paper aims to develop a formal method to model high level interaction between autonomous vehicles. We introduce a concept of road graph to represent complex road situations such as intersections, road merging, unmarked roads, and traffic hazards. We then extend the concept to further represent status of vehicles, dynamics of traffic and protocols of traffic control. Specifically, we formalise two categories of traffic control protocols, time-based protocols and priority-based protocols.