摘要
针对自动驾驶汽车仿真测试场景生成方法进行了系统梳理,综述了自动驾驶汽车仿真测试场景定义、场景解构、数据驱动式场景生成、机理建模式场景生成等方向的最新研究进展,并对测试场景的相关评价与应用进行了总结,最后,提出了未来研究应重点考虑融入中国驾驶场景特征,深化边缘场景生成策略,加速场景构建的标准体系化建设工作。
This paper systematically sorts out the generation methods of simulation test scenarios for autonomous vehicle,summarizes the latest research progress in the fields of autonomous vehicle simulation test scenario definition,scenario deconstruction,scenario generation based on data driven,and scenario generation based on mechanism modeling,and summarizes the relevant evaluation and application of test scenarios.Finally,the paper proposes that future research should focus on integrating the characteristics of Chinese driving scenarios,deepening the research on edge scenario generation strategies,and accelerating the construction of the standard system of scenario construction.
作者
李伟男
王宇
李林润
孟祥哲
王超
刘迪
Li Weinan;Wang Yu;Li Linrun;Meng Xiangzhe;Wang Chao;Liu Di(Global R&D Center,China FAW Corporation Limited,Changchun 130013;National Key Laboratory of Advanced Vehicle Integration and Control,Changchun 130013)
出处
《汽车工程师》
2024年第7期1-10,共10页
Automotive Engineer
关键词
自动驾驶
测试场景
数据驱动
机理建模
Autonomous driving
Test scenario
Data driven
Mechanism modeling