Scholarlink

<p>In Conditional Driving Automation (SAE Level 3, L3), the automated driving system may not be able to respond effectively in the face of a variety of complex situations, and the driver needs to actively intervene and control the vehicle according to the takeover request (TOR), and at this time, the safety of driver takeover is influenced by various factors such as emotions, mental workload, and driving style. In this paper, we investigate how the driver state affects the takeover quality in the case where the L3 system cannot respond effectively in complex traffic environments and thus needs to be taken over by the driver. Through experimental simulation, we collected physiological data including ECG and EEG as well as vehicle operation data, based on which a structural equation model (SEM) was constructed to investigate the complex relationship between driver's emotion, mental workload, driving style and takeover behavior. In addition, we propose a driver takeover risk calculation method for high-speed road takeover scenarios based on the driving risk field theory, and use machine learning algorithms to predict the takeover risk based on the driver's physiological characteristics before the takeover, which will provide a better basis for the design of safety aids for automated driving.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Risk assessment and prediction of conditional driving automation takeover: considering the impact of driver state

  • 期刊论文

  • 发表日期：

    2025年6月2日

查看全文
  • Haochun Huang,
  • Jing Huang,
  • Zehao Zhou,
  • Xiaofang Long

摘要

In Conditional Driving Automation (SAE Level 3, L3), the automated driving system may not be able to respond effectively in the face of a variety of complex situations, and the driver needs to actively intervene and control the vehicle according to the takeover request (TOR), and at this time, the safety of driver takeover is influenced by various factors such as emotions, mental workload, and driving style. In this paper, we investigate how the driver state affects the takeover quality in the case where the L3 system cannot respond effectively in complex traffic environments and thus needs to be taken over by the driver. Through experimental simulation, we collected physiological data including ECG and EEG as well as vehicle operation data, based on which a structural equation model (SEM) was constructed to investigate the complex relationship between driver's emotion, mental workload, driving style and takeover behavior. In addition, we propose a driver takeover risk calculation method for high-speed road takeover scenarios based on the driving risk field theory, and use machine learning algorithms to predict the takeover risk based on the driver's physiological characteristics before the takeover, which will provide a better basis for the design of safety aids for automated driving.