A Hybrid Model of Traffic Assignment and Control for Autonomous Vehicles

Jianglin Qiao, Dave de Jonge, Dongmo Zhang, Carles Sierra, Simeon Simoff

Abstract: This paper proposes a multi-agent based method to describe traffic control optimization for autonomous vehicle assignment problems on road networks. We first present a formal model for abstract road networks. We then extend the road network model into a game-theoretical model based on population games to describe the behavior of autonomous vehicles under intelligent traffic control. Based on this model, we investigate a traffic control optimization problem that aims to improve the efficiency of road networks and provides an algorithm to find an approximate solution. Lastly, our algorithm significantly reduces the total delay of the road network, as demonstrated by the results of our experiments with the Aimsun ( https://www.aimsun.com ) simulation software.