Traffic conflict technology has become an important means to evaluate traffic safety due to the lack of historical accident data. Aiming at the problems of complex traffic flow and frequent accidents in the expressway weaving areas, this paper proposes a new method that uses historical trajectory data of target vehicles to predict whether traffic conflicts will occur in the future, thereby meeting the need for conflict warning. The core of the method includes a trajectory prediction approach based on long short-term memory with attention mechanism (ATT-LSTM) and a traffic conflict prediction method based on predicted trajectories and Time Difference to Collision (TDTC). The experimental results show that the method proposed in this paper performs excellently in rear-end conflict prediction, which can provide a reference basis for the design of vehicle warning devices.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    A Traffic Conflict Prediction Method in Expressway Weaving Areas Based on ATT-LSTM Trajectory Prediction


    Contributors:
    Yu, Dexin (author) / Yang, Yu (author) / Peng, Wanli (author) / Wu, Xincheng (author)


    Publication date :

    2024-06-14


    Size :

    761117 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Expressway Short-Term Traffic Flow Prediction Based on CNN-LSTM

    Ye, Ting / Zou, Fumin / Guo, Feng | Springer Verlag | 2024


    Dissipative Structure Properties of Traffic Flow in Expressway Weaving Areas

    Weihua ZHANG / Lijun XIONG / Qingtong JI et al. | DOAJ | 2024

    Free access


    Expressway Traffic Volume Prediction Using Floating Car Trajectory Data

    Yuan, Xianghong / Yan, Xiaoxiao / Li, Yang | British Library Conference Proceedings | 2019


    Expressway risk prediction method based on LSTM and BF

    XIONG XIAOXIA / LIU QINGCHAO / SHEN YUJIE et al. | European Patent Office | 2021

    Free access