K.L. Xia, Y. Li, N. Zhang, M. Torres

Pages: 113-124

Abstract
In view of the current situation awareness and early warning model of vehicle collision risk at traffic intersection does not consider the discrimination threshold, which leads to low early warning accuracy and high early warning missing rate, a situation awareness and early warning model of vehicle collision risk based on Internet of vehicles technology is proposed. Using the wireless communication technology of Internet of vehicles, the traffic data information is transmitted to the cloud platform to extract the vehicle running state; using the situation awareness theory, the situation awareness framework of vehicle collision risk is established, and the situation rules of vehicle collision risk are extracted by analyzing the observation data of safe driving state. Through the attribute reduction algorithm, the redundant information and standard attributes which are less related to the decision attributes in the universe space are removed, and the discrimination threshold is determined; the vehicle position information is obtained by using the GPS positioning module of the Internet of vehicles, and the collision position and collision time difference are calculated through different vehicle trajectories, so as to realize the situation awareness and early warning of vehicle collision risk at traffic intersections. The experimental results show that the accuracy of the proposed method is 92%, and the missing rate is only 0.2%.
Keywords: internet of vehicles technology; situation awareness theory; traffic intersection vehicle collision; dangerous situation awareness; collision warning