ECSAS: Exploring critical scenarios from action sequence in autonomous driving

ECSAS : Exploring critical scenarios from action sequence in autonomous driving

Kang, Shuting (author): University of Chinese Academy of Sciences, Beijing, China; Institute of Software Chinese Academy of Sciences, Beijing, China

Guo, Heng (author): University of Chinese Academy of Sciences, Beijing, China Institute of Software Chinese Academy of Sciences, Beijing, China;

Zhang, Lijun (author): Institute of Software Chinese Academy of Sciences, Beijing, China

Liu, Guangzhen (author): Institute of Software Chinese Academy of Sciences, Beijing, China

Xue, Yunzhi (author): Institute of Software Chinese Academy of Sciences, Beijing, China

Wu, Yanjun (author): Institute of Software Chinese Academy of Sciences, Beijing, China

(creator_code:org_t)

Institute of Electrical and Electronics Engineers (IEEE), 2023
2023
English.
In: Proceeding of 2023 IEEE 32nd Asian Test Symposium (ATS). - : Institute of Electrical and Electronics Engineers (IEEE). ; , s. 1-6

Related links:: https://urn.kb.se/re...; show more...; https://doi.org/10.1...; show less...

Abstract Subject headings

Rare critical scenarios are crucial to verify the performance of autonomous driving in different situations. Critical scenario generation requires the ability of sampling critical combinations from an infinite parameter space in the logical scenario. Existing solutions aim to explore the correlation of action parameters in the initial scenario rather than action sequences. How to model action sequences so that one can further consider the effects of different action parameters is the bottleneck of the problem. In this paper, we solve the problem by proposing the ECSAS framework. Specifically, we first propose a description language, BTScenario, allowing us to model action sequences of scenarios. We then use reinforcement learning to search for combinations of critical action parameters. Several optimizations are proposed to increase efficiency, including action mask and replay buffer. Experimental results show that our model with strong collision ability and effectively outperforms the existing methods on various nontrivial scenarios.

By the author/editor: Kang, Shuting; Guo, Heng; Su, Peng; Zhang, Lijun; Liu, Guangzhen; Xue, Yunzhi; show more...; Wu, Yanjun; show less...

Kungliga biblioteket hanterar dina personuppgifter i enlighet med EU:s dataskyddsförordning (2018), GDPR. Läs mer om hur det funkar här.
Så här hanterar KB dina uppgifter vid användning av denna tjänst.