Abstract: Based on the analysis of the applicability of three different autonomous emergency braking (AEB) control models in different crash scenarios, and the China in-depth accident study (CIDAS) China traffic accident in-depth research database, PreScan software is used to build the accident scene, and Simulink is connected to add the vehicle control model, and the real collision scene is tested by reconstructing the CIDAS traffic accident database. The simulation results verify that the autonomous emergency braking (AEB) control strategy based on the safety distance model can effectively avoid rear-end collision accidents or mitigate the harm caused by collisions, among which the Mazda model algorithm is relatively conservative and suitable for high-speed scenarios, and the Honda model algorithm and Berkeley model algorithm are relatively radical and suitable for medium and low-speed scenarios, which provides a reference for the evaluation and improvement of active safety technology.

Key words: Autonomous emergency braking; Safe distance model; Braking critical distance; Test and evaluation; Scene construction

摘要： 针对三种不同自动紧急制动（AEB）控制模型在不同碰撞事故场景下的适用性进行分 析，基于中国交通事故深入研究（CIDAS）数据库，利用 PreScan 软件搭建事故场景，并联 通 Simulink 添加车辆控制模型，通过重建 CIDAS 交通事故数据库中真实碰撞事故场景进行测 试。仿真结果验证了基于安全距离模型的 AEB 控制策略可以有效避免追尾碰撞事故或减轻碰 撞造成的危害，其中 Mazda 模型算法相对保守，适用于高速场景，Honda 模型算法和 Berkeley 模型算法相对激进，适用于中低速场景，为主动安全技术的测评与改进提供了参考。

关键词: 自动紧急制动；安全距离模型；制动临界距离；测试评价；场景构建