Indicator of Safety and Comfort Performance: Quantifying Different Driving Styles for Automated Vehicles
Indicator of Safety and Comfort Performance: Quantifying Different Driving Styles for Automated Vehicles
复制标题
DOI：
10.1109/itsc55140.2022.9922265
复制
发表时间：
2022-10
期刊：
影响因子：
--
通讯作者：
Shuo Cheng;Zheng Wang;Bo Yang;Kimihiko Nakano
中科院分区：
文献类型：
--
作者：
Shuo Cheng;Zheng Wang;Bo Yang;Kimihiko Nakano
驾驶性能的定量评估和不同驾驶风格的分析对于自动驾驶汽车（AV）的大规模使用至关重要。因此，本文提出了一种综合指标来量化自动驾驶汽车在动态安全性和舒适性方面的驾驶性能，然后根据该指标对不同的驾驶方式进行定量分类。所有动态信号均基于车辆全局动力学模型进行分析，并通过数学方法制定严格的性能指标。此外，我们还研究了我们提出的指标的分布，以揭示不同的驾驶风格，包括正常、激进和保守驾驶。实验结果表明，所提出的指标可以有效地评估驾驶性能，并可用于对不同驾驶风格进行分类。这些发现表明我们的研究具有巨大的潜力，可以为自动驾驶汽车的规划和控制方法的设计提供基础。
Quantitative assessment of driving performances and analysis of different driving styles are significantly critical for the mass employment of automated vehicles (AVs). There-fore, this paper presents a synthetical indicator to quantify AV s' driving performance with regard to dynamic safety and comfort and then classifies quantitatively different driving styles based on the proposed indicator. All dynamic signals are analyzed based on vehicle global dynamics modeling, and we develop rigorous performance indicators mathematically. Furthermore, the distribution of our proposed indicator is investigated to reveal different driving styles including normal, aggressive, and conservative driving. Experimental results illustrate the proposed indicator can efficiently evaluate driving performance and can be used to classify different driving styles. These findings suggest our research has great potential to provide a foundation for the design of AVs' planning and control methods.