The invention provides a self-adaptive fusion learning automatic driving safety decision-making method oriented to extremely difficult cases. The method comprises the following steps: collecting scene data of an extreme difficult case through a sensor, preprocessing the scene data of the extreme difficult case, and performing feature extraction and selection on the preprocessed scene data; training and optimizing an extreme difficult case scene model by using the classified and labeled extreme difficult case feature data to obtain a trained extreme difficult case scene model; and carrying out safety decision-making on the extreme difficult case dangerous scene by utilizing the trained extreme difficult case scene model. According to the method, through the steps of real-time sensing and extreme difficult case scene data collection, data preprocessing and feature classification extraction, model training and optimization, decision generation, real-time monitoring and feedback and the like, safe and intelligent response to extreme difficult cases is achieved. The possibility and risk of various extremely difficult cases can be analyzed and predicted, and a security decision is generated and optimized.

    本发明提供了一种面向极端难例的自适应融合学习自动驾驶安全决策方法。该方法包括:通过传感器采集极端难例的场景数据,对极端难例的场景数据进行预处理,对预处理后的场景数据进行特征提取和选择:基于深度自编码卷积神经网络构建极端难例场景模型,利用分类标注的极端难例特征数据对极端难例场景模型进行训练和优化,得到训练好的极端难例场景模型;利用训练好的极端难例场景模型进行极端难例危险场景的安全决策。本发明方法通过实时感知和极端难例场景数据收集、数据预处理和特征分类提取、模型训练和优化、决策生成和实时监控与反馈等步骤,实现了对极端难例的安全智能响应。能够分析和预测各种极端难例的可能性和风险,生成并优化安全决策。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Self-adaptive fusion learning automatic driving safety decision-making method for extreme difficult cases


    Weitere Titelangaben:

    面向极端难例的自适应融合学习自动驾驶安全决策方法


    Beteiligte:
    LIANG CI (Autor:in) / WANG RUI (Autor:in) / ZHENG WEI (Autor:in) / ZHAO MINGYANG (Autor:in)

    Erscheinungsdatum :

    2023-09-19


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    IPC:    G06Q Datenverarbeitungssysteme oder -verfahren, besonders angepasst an verwaltungstechnische, geschäftliche, finanzielle oder betriebswirtschaftliche Zwecke, sowie an geschäftsbezogene Überwachungs- oder Voraussagezwecke , DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES / 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



    Automatic driving decision-making safety system and method

    YANG SHICHUN / LU JIAYI / CAO YAOGUANG et al. | Europäisches Patentamt | 2022

    Freier Zugriff

    Interconnection automatic driving decision-making method based on collaborative awareness and adaptive information fusion

    YU QUANLIN / XIA YUYANG / LIU SHUNCHENG et al. | Europäisches Patentamt | 2024

    Freier Zugriff

    Automatic driving decision-making method based on deep reinforcement learning

    LIU CHENGQI / LIU SHAOWEIHUA / ZHANG YUJIE et al. | Europäisches Patentamt | 2024

    Freier Zugriff

    Semantic Traffic Law Adaptive Decision-Making for Self-Driving Vehicles

    Liu, Jiaxin / Wang, Hong / Cao, Zhong et al. | IEEE | 2023


    Automatic driving decision-making method based on rule-assisted reinforcement learning

    ZHENG KAI / SU HAN / ZENG XIMU | Europäisches Patentamt | 2023

    Freier Zugriff