부산대학교 도서관

Identifying traffic platoons and managing vehicles on the road effectively is a challenging task that is currently under investigation both in academia and the industry. The challenges include the need for fast, real-time gathering of relevant information, such as the vehicle’s location, moving direction, and their speed, and then instructing the vehicles in real time to respect the traffic policies according to the gathered information. In this work, we present new algorithms to define platoons that are updated dynamically on the fly, allowing much better control over the traffic, thus improving efficiency. The platoon’s representative vehicle is chosen and the set of vehicles in the platoon is identified based on inductive distance criteria, which are continuously checked, and by considering the road graph topology as well as updating the platoon memberships. In the first part of this paper, we present the main algorithms to identify and control the platoon and demonstrate this detection using a vehicle simulator. We show various policies to control the platoon movement, when detecting obstacles, or other unusual situations. In the second part, we explain how to make vehicles travel roughly in the same direction as the platoon, and join the platoon at particular points or junctions. The described algorithms are demonstrated by extensive simulations showing the platoon behavior over time.