Abstract: Intelligent vehicles improve road traffic safety, but they also introduce new safety hazards. The article analyzes the collision of a typical level 2 automated intelligent vehicle using event data recorder (EDR) and driving data. It is found that the driver had been using the combined driver assistance system for an extended period while experiencing fatigue driving. Despite repeated reminders from the system, the driver is unable to take over the vehicle and eventually the automated system downgraded and exited. The accident occurred when the vehicle deviated from the lane. Based on this case, the architecture diagram of the autonomous driving system and the logic diagram of the human-machine interaction mechanism are established, the importance of warning prompts and minimal risk strategies in guaranteeing the operational safety of intelligent vehicles is explored, and the deficiencies of the current human-machine interaction mechanism of intelligent vehicles are summarized and suggestions for improvement are put forward.

Key words: Intelligent vehicle; Human-machine interaction; Traffic accidents; Case study; Accident investigation

摘要： 智能汽车在改善道路交通安全的同时，也带来新的安全隐患。文章分析了一起典型 2 级自动化智能汽车碰撞事故的事件数据记录系统（EDR）和行车数据，发现驾驶员在疲劳驾 驶状态下长时间使用组合驾驶辅助系统，且在系统反复提醒下依然无法被唤醒并接管车辆， 最后自动化系统降级退出，偏离车道发生事故。基于该案例建立了自动驾驶系统架构图和人 机交互机制逻辑图，探讨了警告提示和最小风险策略在保障智能汽车运行安全中的重要性， 总结了目前智能汽车的人机交互机制存在的不足并提出改进建议。

关键词: 智能汽车；人机交互；交通事故；案例分析；事故调查