Scholarlink

<p>In complex traffic scenarios, traditional fixed-threshold collision avoidance strategies lead to humancomputer conflicts as they fail to account for the heterogeneity in driver behavior and psychological perception. This study proposes a cooperative collision avoidance strategy that integrates quantitative analysis of driving patterns with psychometric models (Stevens’ power law) to optimize personalized driving needs and system reliability. This paper studies NGSIM I-80 highway trajectory data, establishing a driving behavior classification model via K-means clustering and an improved support vector machine (SVM) fusion algorithm to obtain relative vehicle distances. Subsequently, the critical safe distance is combined with Stevens’ law of psychological quantities to propose an autonomous safe collision avoidance framework that fuses driving characteristics to enhance safety and comfort. Compared with existing aggressive braking strategies, the automatic emergency braking optimization model improves effective braking distance by 31.05%, 23.81% and 39.44% under three typical working conditions. This strategy, while maintaining safety, enhances the smoothness and naturalness of the driving process and can simulate individual driving preferences. This method introduces driver psychological characteristics into autonomous driving control, improving the system’s human-machine adaptability and personalization. Psychological modeling is significant for enhancing user trust and alleviating anxiety, providing theoretical and practical support for human centered intelligent driving technology.</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

Research on Intelligent Vehicle Obstacle Avoidance Strategies Integrating Driving Styles and Psychology: Dual Enhancement of Safety and Comfort via Dynamic Safety Distance Modeling

  • 期刊论文

  • 发表日期：

    2026年1月26日

查看全文
  • Tongying Guo,
  • Yan Zheng,
  • Shuo Fan

摘要

In complex traffic scenarios, traditional fixed-threshold collision avoidance strategies lead to humancomputer conflicts as they fail to account for the heterogeneity in driver behavior and psychological perception. This study proposes a cooperative collision avoidance strategy that integrates quantitative analysis of driving patterns with psychometric models (Stevens’ power law) to optimize personalized driving needs and system reliability. This paper studies NGSIM I-80 highway trajectory data, establishing a driving behavior classification model via K-means clustering and an improved support vector machine (SVM) fusion algorithm to obtain relative vehicle distances. Subsequently, the critical safe distance is combined with Stevens’ law of psychological quantities to propose an autonomous safe collision avoidance framework that fuses driving characteristics to enhance safety and comfort. Compared with existing aggressive braking strategies, the automatic emergency braking optimization model improves effective braking distance by 31.05%, 23.81% and 39.44% under three typical working conditions. This strategy, while maintaining safety, enhances the smoothness and naturalness of the driving process and can simulate individual driving preferences. This method introduces driver psychological characteristics into autonomous driving control, improving the system’s human-machine adaptability and personalization. Psychological modeling is significant for enhancing user trust and alleviating anxiety, providing theoretical and practical support for human centered intelligent driving technology.