A distinctive feature of intelligent transportation systems is that vehicles with communication capabilities are able to exchange information with other connected vehicles (CVs), It is also possible for CVs to receive Signal Phase and Timing (SPaT) information at a signalized intersection. Such newly available information offers a great opportunity for predicting future traffic conditions. Traffic prediction has a significant amount of potential in many applications, such as eco-driving, traffic signal control, traffic safety enhancement, among many others. With SPaT information becoming available, prediction is expected to achieve a higher degree of accuracy, particularly in the vicinity of intersections. In this article, we propose a new method for vehicle speed prediction in the next several seconds. The traffic prediction framework is developed based on the well-known second-order Payne-Whitham (PW) model, which is capable of handling mixed traffic in the presence of CVs and legacy vehicles (LVs). By modifying the equilibrium speed appearing in the PW model, it is possible to capture the impact of traffic lights on vehicular flow, resulting in significant improvements on traffic prediction, especially when vehicles approach the intersection. The CVs provide partial measurements of the traffic states, whilst the unknown traffic states are estimated using an unscented Kalman filter (UKF). Future traffic states are obtained by propagating the PW model forward in time. The proposed prediction method is carefully evaluated with real-world traffic data collected on Hwy 55 in Minnesota. Numerical results show that traffic prediction errors are reduced by up to 41.19% with appropriate modifications to the equilibrium speed term of the PW model.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Traffic Prediction for Connected Vehicles on a Signalized Arterial


    Contributors:
    Sun, Wenbo (author) / Wang, Shian (author) / Shao, Yunli (author) / Sun, Zongxuan (author) / Levin, Michael W. (author)


    Publication date :

    2021-09-19


    Size :

    3023940 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English