Abstract: In the environments of highways and express lanes, the L2 intelligent vehicle automated driving assistance system (ADAS) faces more stringent safety challenges. Once there is a situation where the driver's attention is distracted or there are blind spots in the system's functional coverage, it is highly likely to trigger serious traffic accidents. This paper focuses on an in-depth analysis of a typical accident in which a small passenger car, while driving at a high speed with the active cruise control and automatic steering assistance functions activated, collides severely with a fixed retaining wall ahead. During the research process, multi-source data such as the event data recorder system and in-vehicle logs are comprehensively compared. Moreover, closely combining with the working principles of the autonomous driving system in aspects such as perception and recognition, decisionmaking and planning, and human-machine interaction, the key causes for the vehicle's transition from smooth driving to high-speed collision within a short period are systematically and comprehensively sorted out. This study reveals numerous challenging problems that urgently need to be addressed in the human-machine interaction aspect of the current L2 intelligent vehicle ADAS, providing a theoretical basis and practical guidance for enhancing the safety and reliability of the intelligent driving assistance system

Key words: combined driving assistance; automated driving system; human-machine interaction; high-speed collision; in-depth accident analysis

摘要： 在高速公路和快速通道环境下，L2 级智能汽车自动化驾驶辅助系统（ADAS）面临着 更为严苛的安全挑战。一旦出现驾驶员注意力分散，或是系统功能覆盖存在盲区的情况，极 易引发严重的交通事故。文章聚焦于一起小型普通客车在高速行驶时开启主动巡航控制和自 动辅助转向功能，却与前方固定挡墙发生严重撞击的典型事故展开深度剖析。研究过程中， 综合比对了事件数据记录系统、车载日志等多源数据，并紧密结合自动驾驶系统在感知识别、 决策规划以及人机交互等方面的工作原理，系统且全面地梳理出车辆从平稳行驶到在短时间 内高速撞击的关键致因。揭示了当前 L2 级智能汽车 ADAS 在人机交互层面存在的诸多亟待 攻克的难题，为提升智能驾驶辅助系统安全性与可靠性提供了理论依据与实践指导。

关键词: 组合驾驶辅助；自动驾驶系统；人机交互；高速碰撞；事故深度分析