A spatial short-term traffic flow prediction method based on the macroscopic urban traffic network (UTN) model is described and compared to the traditional time series forecasting methods. This paper presents a general macroscopic UTN model by adopting the transfer mechanism of vehicles between road links to represent the future distribution of vehicles in the whole network. Based on the model, we predict the short-term traffic flux without using any historical traffic data, which is completely different from previous approaches. Furthermore, to verify the effectivity of the UTN-based prediction model, we compare it to four classic models including two parametric and two nonparametric methods with the data produced by CORSIM, a commonly used microscopic traffic simulation software. Finally, the comparative results illustrate that the proposed method can reach the level of classic methods and predict the short-term traffic flow timely and accurately both for the steady or suddenly changed traffic states.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Comparison of urban traffic prediction methods between UTN-based spatial model and time series models


    Contributors:
    Xu, Yanyan (author) / Kong, Qing-Jie (author) / Liu, Yuncai (author)


    Publication date :

    2012-09-01


    Size :

    298938 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Comparison between urban traffic models

    Biros, J.F. / Desforges, O. / Ducornet, D. et al. | Tema Archive | 1979


    Graph time-series mixture models for air traffic prediction

    Vardaro, Anthony / Doan, Cuong Thai / Chandra, Kavitha et al. | IEEE | 2013

    Free access

    Urban traffic flow prediction method based on road network spatial heterogeneity

    LI XUAN / HE MUYANG / SU LI et al. | European Patent Office | 2024

    Free access

    Urban Traffic Flow Forecasting Based on Memory Time-Series Network

    Zhao, Shengjian / Lin, Shu / Li, Yunlong et al. | IEEE | 2020