ODD and Behavior Based Scenario Generation for Automated Driving Systems

被引:3
作者
Zhang, Xizhe [1 ]
Khastgir, Siddartha [1 ]
Tiele, Justin-Kiyoshi [2 ]
Takenaka, Kazuhito [2 ]
Hayakawa, Tasuku [3 ]
Jennings, Paul [1 ]
机构
[1] Univ Warwick, WMG, Coventry CV4 7AL, Warwick, England
[2] DENSO AUTOMOT Deutschland GmbH, Corp R&D Unit, D-85386 Eching, Germany
[3] DENSO Corp, Global R&D Tokyo Haneda, Ota Ku, Tokyo 1440041, Japan
基金
英国科研创新办公室;
关键词
Automated driving systems (ADSs); scenario-based testing; safety; safety assurance; V&V; operational design domain (ODD); behavior; scenario generation; coverage;
D O I
10.1109/ACCESS.2024.3350512
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Automated Driving Systems (ADSs) bring many benefits, however their safety assurance poses challenges. Scenario-based testing has been proposed. To provide strong safety evidence for their safety assurance, the scenario-based testing process needs to consider the Operational Design Domain (ODD) of the systems, ODD combined with the behavioural elements can provide a foundation for the scenario generation workflow due to common domain elements between ODDs and scenarios. Based on such background, this paper introduces a novel framework to generate scenarios specifically target on testing the system's claimed ODD. It includes the process going from the system's ODD and behaviour competency, to logical scenarios generation utilising scenario construct rule sets, and to the concretisation of the logical scenarios into concrete scenarios. This paper also draws link towards the part II of the paper series, which illustrates a novel approach for scenario coverage analysis.
引用
收藏
页码:10652 / 10663
页数:12
相关论文
共 39 条
[1]  
Agrawal R., 1993, SIGMOD Record, V22, P207, DOI 10.1145/170036.170072
[2]  
Anastasio A., 2022, P DRIV SIM C EUR VR, P65
[3]  
[Anonymous], Safety Pool Database
[4]  
[Anonymous], 2021, Automated Lane Keeping Systems (ALKS),
[5]  
[Anonymous], 2021, Standard ISO 22737
[6]  
[Anonymous], 2022, ISO 21448:2022
[7]  
[Anonymous], 2023, ISO 34503
[8]  
[Anonymous], 2022, HIGHWAY CODE
[9]  
ASAM e.V, 2021, ASAM OpenXOntology Concept
[10]  
ASAM e.V, 2021, OpenLABEL V1.0.0