S.L. Wu, F.J. Wang, L. Wang, Z.X. Li

Pages: 15-26

Abstract
The traffic flow model is a complex dynamic problem, which has the characteristics of high-dimensional, time-varying, nonlinear and so on. Therefore, the mathematical model established by the traditional theory cannot accurately describe the characteristics of the traffic flow, and cannot accurately identify the traffic state. This paper studies the recognition of urban road traffic status in the context of the Internet of Vehicles. First of all, the V2X data communication scheme is designed to realize the transmission and sharing of vehicle and road information through the use of V2X technology; Then, combining the probability density of traffic flow distribution and the characteristic distribution of traffic flow, the characteristics of road traffic state are extracted; Finally, based on the traffic state characteristics, a hidden Markov model of traffic state characteristics is established, and the traffic state identification is realized by finding the optimal state sequence to reflect the traffic state in the current time window. The experiment shows that compared with the traditional traffic state recognition method, this method has more accurate recognition accuracy and shorter recognition time.
Keywords: internet of vehicles; urban road; traffic status; data sampling; characteristic analysis; status identification