Adversarial Evaluation of Autonomous Vehicles in Lane-Change Scenarios
Autonomous vehicles must be comprehensively evaluated before deployed in cities and highways. However, most existing evaluation approaches for autonomous vehicles are static and lack adaptability, so they are usually inefficient in generating challenging scenarios for tested vehicles. In this paper, the authors propose an adaptive evaluation framework to efficiently evaluate autonomous vehicles in adversarial environments generated by deep reinforcement learning. Considering the multimodal nature of dangerous scenarios, the authors use ensemble models to represent different local optimums for diversity. The authors then utilize a nonparametric Bayesian method to cluster the adversarial policies. The proposed method is validated in a typical lane-change scenario that involves frequent interactions between the ego vehicle and the surrounding vehicles. Results show that the adversarial scenarios generated by the method significantly degrade the performance of the tested vehicles. The authors also illustrate different patterns of generated adversarial environments, which can be used to infer the weaknesses of the tested vehicles.
- Record URL:
-
Availability:
- Find a library where document is available. Order URL: http://worldcat.org/oclc/41297384
-
Supplemental Notes:
- Copyright © 2022, IEEE.
-
Authors:
- Chen, Baiming
- Chen, Xiang
- Wu, Qiong
- Li, Liang
- Publication Date: 2022-8
Language
- English
Media Info
- Media Type: Web
- Features: References;
- Pagination: pp 10333-10342
-
Serial:
- IEEE Transactions on Intelligent Transportation Systems
- Volume: 23
- Issue Number: 8
- Publisher: Institute of Electrical and Electronics Engineers (IEEE)
- ISSN: 1524-9050
- Serial URL: http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=6979
Subject/Index Terms
- TRT Terms: Autonomous vehicles; Evaluation; Lane changing; Machine learning
- Subject Areas: Highways; Operations and Traffic Management; Planning and Forecasting; Vehicles and Equipment;
Filing Info
- Accession Number: 01857603
- Record Type: Publication
- Files: TRIS
- Created Date: Sep 15 2022 9:20AM