The invention provides an intelligent automobile humanoid lane changing track generation method based on a diffusion model. The intelligent automobile humanoid lane changing track generation method comprises the following steps of 1, obtaining and establishing a multi-scene expert data set; 2, training the diffusion model by adopting the multi-scene expert data set; and 3, generating a humanoid lane changing track by using the diffusion model. According to the method, the multi-scene expert data set for collecting the expert lane changing tracks of multiple scenes is established, and the diversity of the lane changing tracks in the expert data set is improved; the method comprises the following steps: training a diffusion model by using an expert lane changing track of a multi-scene expert data set, and training the diffusion model by using collected real human driver data, so that track information generated by the diffusion model better conforms to human lane changing characteristics and has better human-like characteristics; the lane changing track information generated by the diffusion model trained by the data set is more diversified, and a large amount of reliable, professional and diversified human-like lane changing track data can be obtained while the cost is reduced.

    本发明提供一种基于扩散模型的智能汽车类人换道轨迹生成方法,包括以下步骤:步骤一、获取并建立多场景专家数据集:步骤二、采用多场景专家数据集训练扩散模型:步骤三、用扩散模型生成类人换道轨迹。本发明建立采集了多个场景的专家换道轨迹的多场景专家数据集,提高专家数据集内的换道轨迹的多样性;用多场景专家数据集的专家换道轨迹训练扩散模型,利用采集的真实人类驾驶员的数据训练扩散模型,使得扩散模型生成的轨迹信息更符合人类换道特征,具备更好的类人特性;使得用该数据集训练的扩散模型生成的换道轨迹信息更加多样化,在减少了成本的同时,可以获得大量可靠、专业、多样化的类人换道轨迹数据。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Intelligent automobile human-like lane changing track generation method based on diffusion model


    Weitere Titelangaben:

    一种基于扩散模型的智能汽车类人换道轨迹生成方法


    Beteiligte:
    ZHAO JIAN (Autor:in) / LI SIYUAN (Autor:in) / ZHU BING (Autor:in) / SONG DONGJIAN (Autor:in)

    Erscheinungsdatum :

    2023-03-21


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    IPC:    G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS / G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G06T Bilddatenverarbeitung oder Bilddatenerzeugung allgemein , IMAGE DATA PROCESSING OR GENERATION, IN GENERAL



    Intelligent vehicle track lane changing planning method

    WEI ZHENYA / CHEN WUWEI / DING YUKANG et al. | Europäisches Patentamt | 2020

    Freier Zugriff

    Intelligent automobile lane changing influence range quantification method

    XIONG XIAOXIA / HE YU / CHEN HAN et al. | Europäisches Patentamt | 2024

    Freier Zugriff

    Intelligent driving lane changing track generation method and system and computer storage medium

    GUAN JIE / BI YAMENG / QIAO MIAOMIAO et al. | Europäisches Patentamt | 2022

    Freier Zugriff

    Lane changing track generation method and device and storage medium

    SUN HAO | Europäisches Patentamt | 2023

    Freier Zugriff

    Intelligent driving lane changing track generation method and system, electronic equipment and medium

    XIE TENGYU / WANG BIN / CHENG PENG et al. | Europäisches Patentamt | 2024

    Freier Zugriff