不同认知负荷非驾驶任务下高度自动化驾驶接管绩效预测

doi:10.19562/j.chinasae.qcgc.2023.12.015

Abstract

Abstract:

In highly automated driving， accurate prediction of takeover performance is of great significance to improve the safety of automated driving takeover. Based on the design of driving take-over scenarios under different cognitive load non driving-related tasks （NDRT）， the significance of takeover performance indicators and EEG indicators under different cognitive load NDRT of automated driving is analyzed. Using eSense value and brainwave data of the driver as input， a prediction model of take-over performance based on random forest is constructed to analyze the prediction effect of the model within time windows of 3， 5， 7 and 9 s， and the validity of the model is verified. The results show that there are significant differences in take-over time， maximum lateral acceleration， minimum TTC and driver’s eSense under different loads of NDRT. It is found that the random forest has the best prediction performance within 9 s time window， with the accuracy of 0.94. The prediction accuracy and micro-AUC area of random forest are higher than the results of support vector machine， naive Bayes and logistic regression. The proposed method can effectively predict the take-over performance and provide a theoretical basis for the interaction design between the driver and the autonomous vehicle.

Key words: traffic engineering, takeover performance prediction, random forest, cognitive DNRT, highly automated driving

Yanli Ma,Jun Lu,Jieyu Zhu,Xiaoxue Han. Take-over Performance Prediction Under Different Cognitive Loads of Non-driving Tasks in Highly Automated Driving[J].Automotive Engineering, 2023, 45(12): 2330-2337.

Figures/Tables 15

变量	组别	检验统计	标准误差	标准检验统计	p	$p'$
δ波	0B -1B	0.097	0.062	1.560	0.119	0.713
	0B -2B	0.097	0.062	1.560	0.119	0.713
	0B-无	-0.317	0.062	-5.125	0.000	0.000**
	1B-2B	0.000	0.062	0.000	1.000	1.000
	1B-无	-0.221	0.062	-3.565	0.000	0.002**
	2B-无	-0.221	0.062	-3.565	0.000	0.002**
θ波	0B -1B	0.087	0.062	1.411	0.158	0.949
	0B -2B	0.090	0.062	1.448	0.147	0.885
	0B-无	-0.237	0.062	-3.825	0.000	0.001**
	1B-2B	-0.002	0.062	-0.037	0.970	1.000
	1B-无	-0.149	0.062	-2.414	0.016	0.095
	2B-无	-0.147	0.062	-2.377	0.017	0.105
高α波	0B -1B	0.124	0.062	1.996	0.046	0.275
	0B -2B	0.141	0.062	2.275	0.023	0.138
	0B-无	-0.005	0.062	-0.074	0.941	1.000
	1B-2B	-0.017	0.062	-0.279	0.781	1.000
	1B-无	-0.128	0.062	-2.071	0.038	0.230
	2B-无	-0.145	0.062	-2.349	0.019	0.113
低β波	0B -1B	0.238	0.062	3.844	0.000	0.001**
	0B -2B	0.280	0.062	4.531	0.000	0.000**
	0B-无	-0.217	0.062	-3.510	0.000	0.003**
	1B-2B	-0.043	0.062	-0.687	0.492	1.000
	1B-无	-0.021	0.062	-0.334	0.738	1.000
	2B-无	-0.063	0.062	-1.021	0.307	1.000
低γ波	0B -1B	0.047	0.062	0.761	0.446	1.000
	0B -2B	0.048	0.062	0.771	0.441	1.000
	0B-无	-0.158	0.062	-2.553	0.011	0.064
	1B-2B	-0.095	0.062	-1.532	0.126	0.753
	1B-无	-0.205	0.062	-3.315	0.001	0.006**
	2B-无	-0.110	0.062	-1.783	0.075	0.448

变量	组别	检验统计	标准误差	标准检验统计	p	$p'$
注意力	0B -1B	0.339	0.062	5.469	0.000	0.000**
	0B -2B	0.341	0.062	5.515	0.000	0.000**
	0B-无	-0.256	0.062	-4.132	0.000	0.000**
	1B-2B	-0.003	0.062	-0.046	0.963	1.000
	1B-无	-0.083	0.062	-1.337	0.181	1.000
	2B-无	-0.086	0.062	-1.383	0.167	0.999
放松度	0B -1B	0.341	0.062	5.506	0.000	0.000**
	0B -2B	0.450	0.062	7.270	0.000	0.000**
	0B-无	-0.386	0.062	-6.239	0.000	0.000**
	1B-2B	-0.109	0.062	-1.764	0.078	0.466
	1B-无	-0.045	0.062	-0.733	0.463	1.000
	2B-无	-0.064	0.062	-1.031	0.303	1.000
用脑度	0B -1B	0.013	0.062	0.214	0.831	1.000
	0B -2B	0.149	0.062	2.405	0.016	0.097
	0B-无	-0.276	0.062	-4.457	0.000	0.000**
	1B-2B	-0.162	0.062	-2.618	0.009	0.053
	1B-无	-0.263	0.062	-4.243	0.000	0.000**
	2B-无	-0.425	0.062	-6.861	0.000	0.000**

References 14

1	YOON S H， LEE S C， JI Y G. Modeling takeover time based on non-driving-related task attributes in highly automated driving［J］. Applied Ergonomics， 2021， 92： 103343.
2	LI Q， HOU L， WANG Z， et al. Drivers’ visual-distracted take-over performance model and its application on adaptive adjustment of time budget［J］. Accident Analysis and Prevention， 2021， 154： 106099.
3	AYOUB J， DU N， YANG X J， et al. Predicting driver takeover time in conditionally automated driving［J］. IEEE Transactions on Intelligent Transportation Systems， 2022，23（7）： 9580-9589.
4	GREER R， DEO N， RANGESH A， et al. Safe control transitions： machine vision based observable readiness index and data-driven takeover time prediction［C］. International Conference on The Enhanced Safety of Vehicles （ESV）， Japan， 2023， 4.
5	PAKDAMANIAN E， SHENG S， BAEE S， et al. Deeptake： prediction of driver takeover behavior using multimodal data［C］.Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems， 2021： 1-14.
6	GOLD C， HAPPEE R， BENGLER K. Modeling take-over performance in level 3 conditionally automated vehicles［J］. Accident Analysis & Prevention， 2018， 116： 3-13.
7	BRAUNAGEL C， ROSENSTIE W， KASNECI E. Ready for take-over？ a new driver assistance system for an automated classification of driver take-over readiness［J］. IEEE Intelligent Transportation Systems Magazine， 2017， 9（4）： 10-22.
8	DU N， ZHOU F， PULVER E M， et al. Predicting driver takeover performance in conditionally automated driving［J］. Accident Analysis and Prevention， 2020， 148： 105748.
9	ZHOU F， YANG X J， DE WINTER J C F. Using eye-tracking data to predict situation awareness in real time during takeover transitions in conditionally automated driving［J］. IEEE Transactions on Intelligent Transportation Systems， 2021， 23（3）： 2284-2295.
10	GUPTA S， MISHRA R， CHANG Y H， et al. Modeling driver takeover intention in automated vehicles with attention-based CNN algorithm［C］.Proceedings of the Human Factors and Ergonomics Society Annual Meeting. Sage CA： Los Angeles， CA： SAGE Publications， 2022， 66（1）： 1607-1611.
11	鲁光泉，陈发城，李鹏辉，等. 驾驶人跟车风险接受水平对其接管绩效的影响［J］. 汽车工程， 2021， 43（6）： 808-814.
	LU G Q， CHEN F C， LI P H， et al. Effect of drivers’ acceptance level of car-following risk on the takeover performance ［J］. Automotive Engineering， 2021， 43（6）： 808-814.
12	ZHU J， MA Y， LOU Y. Multi-vehicle interaction safety of connected automated vehicles in merging area： a real-time risk assessment approach［J］. Accident Analysis & Prevention， 2022， 166： 106546.
13	王雪松，徐晓妍. 基于自然驾驶数据的危险事件识别方法［J］. 同济大学学报（自然科学版）， 2019， 48（1）.
	WANG X S， XU X Y. Dangerous event recognition method based on natural driving data ［J］. Journal of Tongji University （Natural Science）， 2019， 48（1）.
14	WANG S， GWIZDKA J， CHAOVALITWONGSE W A. Using wireless EEG signals to assess memory workload in the n-back task［J］. IEEE Transactions on Human-Machine Systems，2015， 46（3）： 424-435.

绩效等级	累积频率
绩效等级	接管时间/s	最大横向加速度/（m·s^-2）	最大纵向减速度/（m·s^-2）	最小TTC/s
不佳	>4.2	>3.72	>1.36	<1.21
良好	（3.5，4.2）	（2.53，3.72）	（0.35，1.36）	（1.21，2.16）
优秀	<3.5	<2.53	<0.35	>2.16

特征类型	特征名称	特征定义	具体特征指标	降维后的特征
eSense特征	注意力	大脑精神专注集中程度	均值、最大值、最小值、标准差	均值、标准差
	放松度	大脑放松程度	均值、最大值、最小值、标准差	均值、标准差
	用脑度	大脑实时工作负荷	均值、最大值、最小值、标准差	均值、标准差
EEG特征	θ波	频域信号能量幅值	均值、最大值、最小值、标准差	最小值、标准差
	高α波	频域信号能量幅值	均值、最大值、最小值、标准差	最小值、标准差
	低β波	频域信号能量幅值	均值、最大值、最小值、标准差	均值、标准差

参数		数值	准确率	Δ准确率
默认参数设置		默认	0.927	0.016
参数寻优	弱学习器迭代次数	191	0.944
	决策树最大深度	18
	最大特征数	7
	叶子节点最少样本数	1
	最大叶子节点数	2
	特征评价标准	gini

时间窗	eSense			EEG			eSense+ EEG
	验证集	测试集		验证集	测试集		验证集	测试集
	准确率	准确率	F1值	准确率	准确率	F1值	准确率	准确率	F1值
3 s	0.66	0.71	0.71	0.50	0.55	0.54	0.78	0.79	0.79
5 s	0.73	0.75	0.74	0.70	0.69	0.69	0.86	0.87	0.87
7 s	0.80	0.84	0.84	0.81	0.84	0.83	0.92	0.93	0.93
9 s	0.84	0.88	0.88	0.85	0.89	0.89	0.93	0.94	0.94

[1]	Yanli Ma, Qin Qin, Fangqi Dong, Yining Lou. Takeover Risk Assessment Model Based on Risk Field Theory Under Different Cognitive Secondary Tasks [J]. Automotive Engineering, 2024, 46(1): 9-17.
[2]	Wenmin Long,Guangquan Lu,Xi Shi,Haitian Tan. Determination Method of Forward Collision Warning Indicator and Threshold Based on Drivers’ Car-Following Characteristics [J]. Automotive Engineering, 2022, 44(9): 1339-1349.
[3]	Datong Qin,Kang Wang,Jihao Feng,Yonggang Liu. Research on Shift Schedule of DCT Vehicle on Flat-Straight Road Based on Driving Data Mining [J]. Automotive Engineering, 2022, 44(11): 1763-1771.
[4]	Baicang Guo,Xianyi Xie,Lisheng Jin,Hui Rong,Yang He,Bingdong Ji. Braking Response Time Prediction Model Based on Multi-dimensional Driving Characteristics [J]. Automotive Engineering, 2021, 43(11): 1683-1692.
[5]	Li Yayong, Cai Yingfeng, Chen Long, Sun Xiaoqiang, He Youguo, Zhang Yunshun. ACC Method Considering Driving State of Front and Rear Vehicles [J]. Automotive Engineering, 2019, 41(8): 865-871.
[6]	Zhu Bing, Li Weinan, Wang Zhen, Zhao Jian, He Rui, Han Jiayi. Identification Strategy of Driving Style Based on Random Forest [J]. , 2019, 41(2): 213-218.

Take-over Performance Prediction Under Different Cognitive Loads of Non-driving Tasks in Highly Automated Driving

RichHTML

PDF (PC)

Like

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 15

References 14

Related Articles 6

Metrics

Comments

Recommended 10