The invention discloses a space-time attention LSTM (Long Short Term Memory) vehicle trajectory prediction method based on position-speed. According to the method, a position-speed space-time attention model architecture is designed to process space-time interaction between vehicles according to the requirements for the real-time performance and accuracy of prediction in the field of network-connected vehicle trajectory prediction. A decoder attention layer is used to cyclically generate an automated vehicle ground future trajectory. By combining location and velocity features, the refined model is used to generate a more accurate vehicle trajectory. Driving information of surrounding vehicles is obtained in real time through a wireless network, input of the prediction module is calculated, and then a vehicle prediction track is obtained. According to the model provided by the method, time attention weight learning can be carried out on vehicle individuals in the whole observation time, and space interaction relations among all vehicles in a grid map area are extracted; in order to further improve the accuracy, results generated by the position and velocity vectors are transmitted to a refinement model for processing.

    本发明公开了一种基于位置‑速度的时空注意力LSTM车辆轨迹预测方法。该方法针对网联车辆轨迹预测领域对预测的实时性和准确性要求,设计了一种位置‑速度的时空注意力模型架构来处理车辆之间的时空交互。解码器注意力层被用来循环生成自动车辆地未来轨迹。通过结合位置和速度特征,细化模型用于生成更准确的车辆轨迹。通过无线网络实时获取周围车辆行驶信息,计算预测模块的输入,进而得到车辆预测轨迹。本发明的方法提出的模型可以在整个观测时间中对车辆个体进行时间注意力权重学习,并提取网格地图区域内所有车辆之间的空间交互关系;为了进一步提高精确度,对位置和速度向量产生的结果传输至细化模型进行处理。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Space-time attention LSTM vehicle trajectory prediction method based on position-speed


    Weitere Titelangaben:

    一种基于位置-速度的时空注意力LSTM车辆轨迹预测方法


    Beteiligte:
    SONG XIULAN (Autor:in) / YANG LUSHENG (Autor:in) / DONG ZHAOHANG (Autor:in) / HE DEFENG (Autor:in) / ZHU WEI (Autor:in)

    Erscheinungsdatum :

    2023-10-13


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    IPC:    B60W CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION , Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion / G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen



    Probabilistic Vehicle Trajectory Prediction Based on LSTM Encoder-Decoder and Attention Mechanism

    Meng, Dejian / Zhang, Lijun / Xiao, Wei et al. | SAE Technical Papers | 2022


    Two-stream LSTM Network with Hybrid Attention for Vehicle Trajectory Prediction

    Li, Chao / Liu, Zhanwen / Zhang, Jiaying et al. | IEEE | 2022


    Vehicle trajectory prediction based on LSTM network

    Yang, Zhifang / Liu, Dun / Ma, Li | IEEE | 2022


    Transform and LSTM-based vehicle trajectory prediction method

    CHENG DENGYANG / GU XIANG / QIAN CONG et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    Probabilistic Vehicle Trajectory Prediction Based on LSTM Encoder-Decoder and Attention Mechanism

    Zhang, Lijun / Liu, Zihao / Xiao, Wei et al. | British Library Conference Proceedings | 2022