The invention discloses a multi-scene, multi-agent cooperation and game-oriented trajectory prediction method. The method comprises the following steps: acquiring trajectory planning information of a plurality of agents; acquiring historical track information of a plurality of intelligent agents and obstacles; acquiring operation scene information of a plurality of intelligent agents and obstacles; and inputting the trajectory planning information, the historical trajectory information and the operation scene information into the trajectory prediction model to obtain prediction trajectories of the intelligent agents and the obstacles. According to the method, the game relationship between multiple agents is considered under the condition of cooperation of the multiple agents and obstacles, the planned trajectory information of the multiple agents is added into the trajectory prediction model, more prior information is provided for trajectory prediction, the robustness of the trajectory prediction model is improved by learning the influence of the diversity of the planned trajectory on the motion trajectory, and the trajectory prediction efficiency is improved. And thus, a more accurate and reasonable prediction track can be obtained.

    本发明公开一种面向多场景、多智能体协同和博弈的轨迹预测方法,包括:获取若干智能体的轨迹规划信息;获取若干智能体及障碍物的历史轨迹信息;获取若干智能体及障碍物的运行场景信息;将轨迹规划信息、历史轨迹信息、运行场景信息输入轨迹预测模型,获取各个智能体及障碍物的预测轨迹。本发明考虑到多智能体、障碍物协同的情况下,多智能体之间存在博弈关系,将多智能体的规划轨迹信息加入轨迹预测模型,为轨迹预测提供更多的先验信息,通过学习规划轨迹的多样性对运动轨迹的影响,提升轨迹预测模型的鲁棒性,从而得到更为准确和合理的预测轨迹。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Trajectory prediction method and device for multi-scene and multi-agent collaboration and gaming


    Weitere Titelangaben:

    面向多场景、多智能体协同和博弈的轨迹预测方法及设备


    Beteiligte:
    HE TAO (Autor:in) / CHEN XIAOLEI (Autor:in) / LIAO WENLONG (Autor:in) / PENG PAI (Autor:in)

    Erscheinungsdatum :

    2023-12-05


    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



    MULTI-AGENT TRAJECTORY PREDICTION

    NARAYANAN SRIRAM NOCHUR / LIU BUYU / MOSLEMI RAMIN et al. | Europäisches Patentamt | 2021

    Freier Zugriff

    Multi-agent trajectory prediction

    NARAYANAN SRIRAM NOCHUR / LIU BUYU / MOSLEMI RAMIN et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    MULTI-MODAL MULTI-AGENT TRAJECTORY PREDICTION

    SUN PEI / ZHAO HANG / MCCAULEY ALEXANDER et al. | Europäisches Patentamt | 2022

    Freier Zugriff

    Multi-scene trajectory prediction method and device, medium and vehicle

    QIN HAIBO / PENG QIXIANG / LI CHUANKANG et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    Multi-agent future trajectory prediction system and method

    YAN CHENGGANG / CHEN HONGLEI / ZHAI CHUNJIE et al. | Europäisches Patentamt | 2023

    Freier Zugriff