L3级自动驾驶系统安全认证方法  

Safety Certification Methods of L3-level Autonomous Driving Systems

作　　者：高雅[1] 朱西产[1] GAO Ya;ZHU Xichan(School of Automotive Studies,Tongji University,Shanghai 201804,China)

机构地区：[1]同济大学汽车学院,上海201804

出　　处：《佳木斯大学学报（自然科学版）》2022年第3期46-49,共4页Journal of Jiamusi University：Natural Science Edition

摘　　要：为了保证L3级自动驾驶汽车上路行驶的安全性,提出了针对L3级自动驾驶系统的安全认证方法框架,主要由构建危险功能场景库、基于具体功能场景确定临界工况和在临界工况下完成封闭道路测试三部分组成。首先介绍了通过拆解场景元素的结构化描述方法来定义自动驾驶系统的危险功能场景,接下来以自动车道保持系统(Automated Lane Keeping System,ALKS)为例介绍了特定危险功能场景下通过构建驾驶员模型获取临界工况的方法,并最终根据相关法规提出的临界工况设计并开展了相关试验,试验结果证明了试验车辆在临界工况下具有较好的避撞能力。To ensure the safety of L3 level autonomous vehicles on the road,a safety certification method framework for L3 level autonomous driving systems is proposed,which consists of three parts:constructing a hazard function scenario library,determining critical conditions based on specific function scenarios,and completing field testing under critical conditions.Firstly,a structured description method by disassembling scenario elements is introduced to define the hazardous functional scenarios of autonomous driving systems.Next,the Automated Lane Keeping System(ALKS)is used as an example to introduce the method to obtain the critical conditions by constructing a driver model for a specific hazard function scenario.Finally,testing was conducted under the critical conditions proposed by the relevant regulations,and the testing results prove that the testing vehicle has good collision avoidance capability under the critical conditions.

关 键 词：自动驾驶 安全认证 测试与验证 交通场景 

分 类 号：U471.15[机械工程—车辆工程]