Crossroad control algorithm based on V2I-interaction

Abstract

The importance of efficient traffic management problem grows every year. Constantly increasing number of vehicles and traffic volume requires improvement of methods and algorithms of traffic control. Such improvement becomes possible due to the spreading of global navigation and positioning systems, the development of wireless communication technologies and mobile internet, the advance in specialized technologies for interaction between vehicles and road infrastructure (V2I), the enhancement of machine learning models, computer vision algorithms, the emergence of driverless cars. This paper considers the novel approach to traffic management at intersection based on the use of V2I communication, describes general scheme of such approach and differences with conventional traffic light regulation. Developed algorithm of intersection management utilizes advantages of V2I communication to increase throughput of the intersection comparing to simple traffic light regulation and more advanced adaptive methods. The increase of throughput is achieved due to the dynamic construction of regulation phases which can be done because of more complete information about traffic flow. The comparence of various methods of regulation was performed in developed traffic simulation environment based on multi-agent approach.