驾驶员接管自动驾驶车辆的眼动特性和行为分析

doi:10.16265/j.cnki.issn1003-3033.2022.01.009

中国安全科学学报 ›› 2022, Vol. 32 ›› Issue (1): 65-71.doi: 10.16265/j.cnki.issn1003-3033.2022.01.009

• 安全社会科学与安全管理 • 上一篇下一篇

驾驶员接管自动驾驶车辆的眼动特性和行为分析

郭子慧¹^,²(), 郭伟伟¹, 谭墍元¹

¹北方工业大学城市道路交通智能控制技术北京市重点实验室,北京 100144
²交通运输部公路科学研究院,北京 100088

收稿日期:2021-10-14 修回日期:2021-12-21 出版日期:2022-01-28 发布日期:2022-07-28
作者简介:
郭子慧(1996—),女,北京人,硕士研究生,研究方向为交通安全和驾驶行为。E-mail: zh.guo@rioh.cn。
郭伟伟副教授,谭墍元副教授
基金资助:
国家自然科学基金资助(61503007); 北京市青年拔尖人才项目(CIT&TCD201804006); 北京市青年拔尖人才项目(CIT&TCD201904013); 中央级公益性科研院所基本科研业务费专项资金项目(2018-9015)

Analysis on eye movement characteristics and behavior of drivers taking over automated vehicles

GUO Zihui¹^,²(), GUO Weiwei¹, TAN Jiyuan¹

¹Beijing Key Lab of Urban Intelligent Traffic Control Technology, North China University of Technology, Beijing 100144, China
²Research Institute of Highway Ministry of Transport, Beijing 100088, China

Received:2021-10-14 Revised:2021-12-21 Online:2022-01-28 Published:2022-07-28

摘要/Abstract

摘要：

为探究接管自动驾驶车辆期间驾驶员的视觉特性,分析眼动与接管反应操控行为的关系,开展驾驶模拟试验收集驾驶行为及眼动数据。运用统计学方法,分析驾驶员感知不同接管场景的视觉特性,探究接管请求(TOR)前后眼动指标的变化规律;并基于视觉分配和瞳孔变化特性分析驾驶行为,揭示眼动特性与接管反应及驾驶操纵策略的内在联系。结果表明:TOR前,相较于静态场景,驾驶员感知动态场景诱发元素扫视更频繁且平均注视时间更短;此时驾驶员的视觉分配特性与其接管反应行为存在显著相关性。TOR后,驾驶员的注视时间增加,眨眼频率降低,瞳孔直径扩张,眼跳幅度增大;不同场景下驾驶员的瞳孔差异表明其应对动态场景时具备更好的警戒水平和更平稳的操纵策略。

关键词: 驾驶员, 接管请求(TOR), 自动驾驶车辆, 眼动特性, 驾驶行为

Abstract:

In order to explore visual characteristics of drivers while taking over automated vehicles, relationship between their eye movement and take-over response and manipulation behavior was analyzed, and driving simulation tests were conducted to collect driving behavior and eye movement data. Then, their visual perception characteristics in different take-over scenarios were analyzed by using statistical method, and variance law of eye movement indexes before and after TOR were explored. Finally, driving behavior was analyzed based on visual distribution and pupil change characteristics, and intrinsic relationship between eye movement features and take-over response and driving manipulation strategies was revealed. The results show that compared with static scenarios, drivers perceive elements induced from dynamic scenarios more frequently and the average fixation time is much shorter before TOR. Meanwhile, their visual distribution characteristics have a significant correlation with take-over behavior. After TOR, drivers' fixation time increases, while blinking frequency decreases, pupil diameter dilates and saccade amplitude increases. The pupil difference of drivers in different scenarios indicates that they have better vigilance level and more stable manipulation strategy when dealing with dynamic scenarios.

Key words: drivers, take-over request(TOR), automated vehicles, eye movement characteristics, driving behavior

郭子慧, 郭伟伟, 谭墍元. 驾驶员接管自动驾驶车辆的眼动特性和行为分析[J]. 中国安全科学学报, 2022, 32(1): 65-71.

GUO Zihui, GUO Weiwei, TAN Jiyuan. Analysis on eye movement characteristics and behavior of drivers taking over automated vehicles[J]. China Safety Science Journal, 2022, 32(1): 65-71.

图/表 7

图1

图2

图3

表1

表2

图4

图5

参考文献 20

[1]	陈清泰. 未来汽车如何与交通、城市协同发展[J]. 汽车纵横, 2020(12):22-23.
[2]	边明远, 李克强. 高级别自动驾驶的技术特征和落地场景[J]. 智能网联汽车, 2019(2):23-27.
[3]	SAE International. Surface vehicle recommended practice: taxonomy and definitions for terms related to driving automation systems for on-road motor vehicles[R], 2018.
[4]	乔建刚, 郭蕊, 李维东. 穿城镇道路驾驶员眼动特性分析[J]. 中国安全科学学报, 2020, 30(7):13-18. doi: 10.16265/j.cnki.issn1003-3033.2020.07.003
	QIAO Jiangang, GUO Rui, LI Weidong. Eye movement characteristics of drivers while crossing urban highways[J]. China Safety Science Journal, 2020, 30(7):13-18.
[5]	汪磊, 高杉. 飞行冒险行为的眼动和生理特征研究[J]. 中国安全科学学报, 2020, 30(9):22-28.
	WANG Lei, GAO Shan. Study on eye movement and physiological characteristics of flying risk-taking behaviors[J]. China Safety Science Journal, 2020, 30(9):22-28.
[6]	VICTOR T W, HARBLUK J L, ENGSTROM J A. Sensitivity of eye movement measures to in-vehicle task difficulty[J]. Transportation Research Part F: Traffic Psychology and Behaviour, 2005, 8(2):167-190. doi: 10.1016/j.trf.2005.04.014
[7]	MERAT N, JAMSON A H, LAI F C H, et al. Highly automated driving, secondary task performance, and driver state[J]. Human Factors the Journal of the Human Factors and Ergonomics Society, 2012, 54(5):762-771. doi: 10.1177/0018720812442087
[8]	SCHWALM M, KEINATH A, ZIMMER H D. Pupillometry as a method for measuring mental workload within a simulated driving task[C]. Proceedings of the Annual Meeting of Human Factors and Ergonomics Society-Europe Chapter, 2008: 1-13.
[9]	ZEEB K, BUCHNER A, SCHRAUF M. What determines the take-over time? an integrated model approach of driver take-over after automated driving[J]. Accident Analysis and Prevention, 2015, 78:212-221. doi: 10.1016/j.aap.2015.02.023
[10]	WU Yanbin, KIHARA K, TAKEDA Y, et al. Effects of scheduled manual driving on drowsiness and response to take over request: a simulator study towards understanding drivers in automated driving[J]. Accident Analysis and Prevention, 2019, 124:202-209. doi: 10.1016/j.aap.2019.01.013
[11]	ZHOU Fang, ALSAID A, BLOMMER M, et al. Driver fatigue transition prediction in highly automated driving using physiological features[J]. Expert Systems with Applications, 2020, 147:DOI:10.1016/j. eswa.2020.113204.
[12]	LOUW T, KUO J, ROMANO R, et al. Engaging in NDRTs affects drivers' responses and glance patterns after silent automation failures[J]. Transportation Research Part F: Traffic Psychology and Behaviour, 2019, 62:870-882. doi: 10.1016/j.trf.2019.03.020
[13]	YOON S H, JI Yonggu. Non-driving-related tasks, workload, and takeover performance in highly automated driving contexts[J]. Transportation Research Part F: Psychology and Behaviour, 2019, 60:620-631. doi: 10.1016/j.trf.2018.11.015
[14]	WINTER J C F D, HAPPEE R, MARTENS M H, et al. Effects of adaptive cruise control and highly automated driving on workload and situation awareness: a review of the empirical evidence[J]. Transportation Research Part F: Psychology and Behaviour, 2014, 27:196-217. doi: 10.1016/j.trf.2014.06.016
[15]	MERAT N, JAMSON A H, LAI F C H, et al. Transition to manual: driver behaviour when resuming control from a highly automated vehicle[J]. Transportation Research Part F: Traffic Psychology and Behaviour, 2014, 27:274-282. doi: 10.1016/j.trf.2014.09.005
[16]	GROH K, KUEHBECK T, FLEISCHMANN B, et al. Towards a scenario-based assessment method for highly automated driving functions[C]. Proceedings of the 8^th Conference on Driver Assistance, 2017:1-7.
[17]	GOLD C, DAMBOCK D, LORENZ L, et al. "Take over!" how long does it take to get the driver back into the loop?[C]. Proceedings of the Human Factors and Ergonomics Society 57^th Annual Meeting, 2013:1938-1942.
[18]	殷莉, 吴玲, 路巧珍, 等. 基于扫视特征的车载信息系统与驾驶安全关系研究[J]. 中国安全科学学报, 2015, 25(6):124-128.
	YIN Li, WU ling, LU Qiaozhen, et al. Research on relationship between in-vehicle information system and driving safety based on scanning features[J]. China Safety Science Journal, 2015, 25(6):124-128.
[19]	焦方通, 杜志刚, 王首硕, 等. 城市水下特长隧道弯道驾驶人扫视行为研究[J]. 中国安全科学学报, 2019, 29(7):104-109. doi: 10.16265/j.cnki.issn1003-3033.2019.07.017
	JIAO Fangtong, DU Zhigang, WANG Shoushuo, et al. Research on saccade behavior of drivers in extra-long urban underwater tunnel curves[J]. China Safety Science Journal, 2019, 29(7):104-109. doi: 10.16265/j.cnki.issn1003-3033.2019.07.017
[20]	商艳, 朱守林, 戚春华, 等. 草原公路景观要素影响驾驶员眼动指标研究[J]. 中国安全科学学报, 2017, 27(5):19-24. doi: 10.16265/j.cnki.issn1003-3033.2017.05.004
	SHANG Yan, ZHU Shoulin, QI Chunhua, et al. Impacts of grassland highway landscape elements on driver's eye movement indices[J]. China Safety Science Journal, 2017, 27(5):19-24. doi: 10.16265/j.cnki.issn1003-3033.2017.05.004

眼动指标	均值(标准差)		t值(显著性)	95%置信区间
眼动指标	TOR前5 s	TOR后5 s	t值(显著性)	95%置信区间
注视持续时间/s	3.42(0.91)	4.14(0.82)	6.167^***	(0.53, 0.91)
扫视持续时间/s	0.51(0.25)	0.44(0.16)	-1.266	(-0.04, 0.17)
眨眼频率/(次·s^-1)	0.56(0.34)	0.37(0.21)	-3.486^**	(-1.02, -0.26)
瞳孔直径/mm	3.54(0.75)	3.89(0.72)	5.670^***	(0.21, 0.43)
眼跳幅度/(°)	2.65(0.45)	4.22(0.81)	7.262^***	(0.88, 1.51)

指标		接管反应时间	制动反应时间	转向反应时间
扫视持续时间	相关系数	-0.340^**	-0.240^*	0.419^***
扫视持续时间	显著性	0.004	0.028	<0.001
注视持续时间	相关系数	0.191	0.146	-0.261^*
注视持续时间	显著性	0.069	0.165	0.016

驾驶员接管自动驾驶车辆的眼动特性和行为分析

Analysis on eye movement characteristics and behavior of drivers taking over automated vehicles

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 7

参考文献 20

相关文章 15

编辑推荐

Metrics

本文评价

[1]	李琼, 汪勇杰, 陈文强, 白鹏霞, 薛晨蕾. 公交驾驶员组织认同感对驾驶愤怒的影响:职业倦怠的中介作用[J]. 中国安全科学学报, 2022, 32(5): 8-13.
[2]	秦雅琴, 李秋谷, 赵鹏燕, 宝发旺, 谢济铭. 基于多分类Adaboost算法的驾驶人风险感知倾向研究[J]. 中国安全科学学报, 2022, 32(4): 141-147.
[3]	赵小平, 马有才, 万平, 莫振龙. 情绪对新手驾驶员视觉特性的影响研究[J]. 中国安全科学学报, 2022, 32(1): 34-40.
[4]	王嘉文, 游权, 周溪召. 城市道路信号控制交叉口两难区研究综述^*[J]. 中国安全科学学报, 2021, 31(8): 125-131.
[5]	韩天园, 李旋, 吕凯光, 周文财, 刘永涛, 乔洁. 车载GPS数据驱动的危货运驾驶员安全绩效评价^*[J]. 中国安全科学学报, 2021, 31(7): 164-171.
[6]	商艳, 朱守林, 戚春华. 草原公路长直线段最佳信息量阈值研究[J]. 中国安全科学学报, 2021, 31(6): 189-195.
[7]	尚婷, 吴鹏, 唐伯明, 白婧荣, 周亮宇. 隧道至互通小间距路段车辆换道博弈行为研究^*[J]. 中国安全科学学报, 2021, 31(10): 68-75.
[8]	朱冠吉, 刘迪, 孙强, 周子超. 铁路机车驾驶员职业安全适应性评价[J]. 中国安全科学学报, 2020, 30(S1): 64-70.
[9]	乔建刚, 郭蕊, 李维东. 穿城镇道路驾驶员眼动特性分析[J]. 中国安全科学学报, 2020, 30(7): 13-18.
[10]	张惠玲, 孔德学, 敖谷昌. 基于不同风险倾向的驾驶员路径诱导模型[J]. 中国安全科学学报, 2020, 30(5): 108-114.
[11]	罗浩顺, 徐良杰, 杜志刚, 曾欢. 城市长隧道导向视觉参照系安全改善效果研究[J]. 中国安全科学学报, 2020, 30(4): 154-159.
[12]	李梦琦, 张敬磊. 考虑愤怒情绪的驾驶员换道行为博弈分析[J]. 中国安全科学学报, 2020, 30(2): 165-170.
[13]	史雪莹, 孙智诚, 朱彤. 基于异质性驾驶行为与碰撞的中观交通流仿真[J]. 中国安全科学学报, 2019, 29(9): 64-69.
[14]	王晓勇, 罗珅, 任杰, 朱彤. 公交驾驶员违规间隔时间及影响因素研究[J]. 中国安全科学学报, 2019, 29(6): 128-133.
[15]	刘玲莉, 田东东, 王子越. 私家车驾驶员驾驶行为与人格特质的关系研究[J]. 中国安全科学学报, 2019, 29(1): 13-18.