Short-term Traffic Flow Prediction in a Partially Connected Vehicle Environment

Short-term traffic flow prediction remains one of the important tasks considered in intelligent transportation systems. Accurate prediction of traffic flow is necessary for traffic management, route guidance, urban planning, and others. One of the current trends in the development of transportation systems is the development of connected and/or autonomous vehicles. Using information from connected vehicles, it is possible to more accurately predict the traffic flow. In this article, we consider the traffic flow in a partially connected vehicle environment, consisting of connected and non-connected (human-driven) vehicles. We propose a short-term mixed traffic flow prediction algorithm that combines an algorithm for traffic flow prediction based on a graph neural network with an algorithm for anticipatory routing of connected vehicles. Experimental studies of the algorithm depending on the penetration rate of connected vehicles were carried out in the microscopic traffic simulation package SUMO. The results confirmed the efficiency of the proposed algorithm.