The invention relates to a vehicle lane change trajectory prediction method based on physical information deep learning, and the method comprises the following steps: S1, constructing a data set, dividing the data set into a training set, a verification set and a test set, extracting an observation state # imgabs0 # and a matching state Ij from the training set, and extracting an observation state # imgabs1 # from the verification set; the method comprises the following steps of: selecting a physical prediction model and an Attention-TCN model, forming a physical information deep learning model with the Attention-TCN model, training, calculating loss functions and loss on a verification set, and comparing the loss on the verification set with an optimal loss value and the loss functions to selectively update parameters of the Attention-TCN model and parameters of the physical prediction model; and S3, performing a test experiment by using the trained physical information deep learning model, and comparing the prediction performance of different lane changing trajectory prediction models. And the constructed physical prediction model and the Attention-TCN model are combined to form a physical information deep learning model, so that the interpretability and descriptiveness of lane changing trajectory prediction are improved.

    本发明涉及一种基于物理信息深度学习的车辆换道轨迹预测方法,包括如下步骤:S1:构建数据集,并将数据集划分为训练集、验证集和测试集,从训练集中提取观测状态#imgabs0#和搭配状态Ij,并从验证集中提取观测状态#imgabs1#S2:构建物理预测模型,并与Attention‑TCN模型组成物理信息深度学习模型,并进行训练,计算损失函数和验证集上的损失,并比较验证集上的损失与最优损失值以及损失函数之间的大小,来选择性对Attention‑TCN模型的参数和物理预测模型的参数进行更新;S3:利用训练后的物理信息深度学习模型进行测试实验,对比不同的换道轨迹预测模型的预测性能。通过将构建的物理预测模型与Attention‑TCN模型结合组成物理信息深度学习模型,提高了换道轨迹预测的可解释性和可描述性。


    Access

    Download


    Export, share and cite



    Title :

    Vehicle lane changing track prediction method based on physical information deep learning


    Additional title:

    一种基于物理信息深度学习的车辆换道轨迹预测方法


    Contributors:
    REN MINGLUN (author) / LI XIXI (author) / WU SHUHUI (author) / CHEN NENGYING (author) / LUO LIJIA (author) / LI YUPING (author) / CHEN ZIXUAN (author) / XU HONGMENG (author)

    Publication date :

    2024-09-20


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    Chinese


    Classification :

    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



    Intelligent vehicle lane changing collision probability evaluation method based on track prediction

    DUAN YUCHEN / DENG YUCONG / QI WEIWEI et al. | European Patent Office | 2021

    Free access

    Intelligent vehicle track lane changing planning method

    WEI ZHENYA / CHEN WUWEI / DING YUKANG et al. | European Patent Office | 2020

    Free access

    Vehicle track deep learning prediction method considering physical constraint

    XU FANG / HAN SHUAI / GUO ZHONGYI et al. | European Patent Office | 2023

    Free access

    Knowledge data combined driving automatic driving vehicle lane changing track prediction method

    SUN CHAO / WEN DA / NING CHANGJIU et al. | European Patent Office | 2024

    Free access

    Driver Lane-Changing Behavior Prediction Based on Deep Learning

    Cheng Wei / Fei Hui / Asad J. Khattak | DOAJ | 2021

    Free access