Research on dangerous scene of T-intersection based on NAIS database accident data clustering

doi:10.3969/j.issn.1674-8484.2021.03.008

Abstract

Abstract:

Based on the National Automobile Accident in-Depth Investigation System (NAIS) database, the T-intersection dangerous scenarios in the traffic environment was investigated. 4 types of T-intersection dangerous accident scenarios were obtained with MATLAB software for cluster analysis and by using the hierarchical clustering method. Through the kinematic analysis and modeling of the 3 types of dangerous scenes when turning, the speed-distance dangerous model was constructed, and the kinematic parameters were combined to obtain the dangerous collision domains under the 3 types of T-junction turning dangerous scenes. The results show that under the conditions of conflict between a passenger vehicle going straight and a passenger vehicle turning left, the dangerous collision range is 1.67~68.45 m when the passenger car turns left at entrance 1; while the dangerous collision range is -1.20~136.00 m when the passenger car turns left at entrance 3; when a commercial vehicle turns right and a two-wheeler goes straight, the dangerous collision range is 0.78~121.30 m; the dangerous collision range is -15.75~283.30 m when a commercial vehicle turns right and a commercial vehicle goes straight. The results of the research can be used as a reference for the development and testing of auxiliary driving safety systems.

Key words: driving safety, traffic accident data, dangerous scene, cluster analysis, kinematics analysis, T-intersection

CLC Number:

X951

LIAO Jingqian, ZHANG Daowen, GAO Li, LIAO Wenjun. Research on dangerous scene of T-intersection based on NAIS database accident data clustering[J]. Journal of Automotive Safety and Energy, 2021, 12(3): 336-345.

Figures/Tables 14

References 25

[1]	国家统计局. 交通事故年度数据[/OL]. [2020-12-20]. https://data.stats.gov.cn/easyquery.htm?cn=C01&zb=A0S0D01&sj=2019.
	National Bureau of Statistics.Annual traffic accident data [/OL]. [2020-12-20], https://data.stats.gov.cn/easyquery.htm?cn=C01&zb=A0S0D01&sj=2019. (in Chinese)
[2]	徐向阳, 周兆辉, 胡文浩, 等. 基于事故数据挖掘的AEB路口测试场景[J/OL]. 北京航空航天大学学报: 1-12. [2020-08-21], https://doi.org/10.13700/j.bh.1001-5965.2019.0548.
	XU Xiangyang, ZHOU Zhaohui, HU Wenhao, et al. AEB intersection test scenario based on accident data mining [J/OL]. *J Beijing Univ Aero Astro*: 1-12. [2020-08-21], https://doi.org/10.13700/j.bh.1001-5965.2019.0548. (in Chinese)
[3]	公安部交通管理局. 中华人民共和国道路交通事故数据统计年报[M]. 无锡:公安部交通管理科学研究所, 2018: 48-49.
	Traffic Administration of the Ministry of Public Security. The People’s Republic of China Road Traffic Accident Data Statistics Annual Report[M]. Wuxi: Traffic Management Research Institute of the Ministry of Public Security, 2018: 48-49. (in Chinese)
[4]	胡林, 易平, 黄晶, 等. 基于真实事故案例的自动紧急制动系统两轮车测试场景研究[J]. 汽车工程, 2018, 40(12):1435-1446+1453.
	HU Lin, YI Ping, HUANG Jing, et al. Research on automatic emergency braking system two-wheeled vehicle test scenario based on real accident cases[J]. *Autom Engi*, 2018, 40(12):1435-1446+1453. (in Chinese)
[5]	范天赐, 王宏雁, 杨震, 等. 基于中国两轮车事故的典型场景提取与分析[J]. 交通与运输, 2019, 35(2):41-45.
	FAN Tianci, WANG Hongyan, YANG Zhen, et al. Extraction and analysis of typical scenes based on two-wheeler accidents in China[J]. *Traff Transport*, 2019, 35(2):41-45. (in Chinese)
[6]	Sui B, Lubbe N, Bargman J. A clustering approach to developing car-to-two-wheeler test scenarios for the assessment of automated emergency braking in China using in-depth Chinese crash data[J]. *Accid Anal Prev, 2019, 132*(1):1-12.
[7]	毛攀. 汽车与两轮车碰撞事故场景分析及AEB避撞研究[D]. 成都:西华大学, 2020.
	MAO Pan. Scene analysis of car-two-wheeler collision accident and AEB collision avoidance research[D]. Chendu: Xihua University, 2020. (in Chinese)
[8]	侯彦巧. 基于典型场景的自动紧急制动系统对骑车人保护效果研究[D]. 成都:西华大学, 2020.
	HOU Yanqiao. Research on the protection effect of automatic emergency braking system on cyclists based on typical scenarios[D]. Chendu: Xihua University, 2020.
[9]	HUANG Sunan, YANG Jikuang, Eklud F. Evaluation of remote pedestrian sensor system based on the analysis of car-pedestrian accident scenarios[J]. *Safe Sci*, 2007, 46(9):1345-1355. doi: 10.1016/j.ssci.2007.08.004 URL
[10]	Good D, Abrahams R. Vehicle speeds for pedestrian pre-crash system test scenarios based on us data [C]// 2014 IEEE 17^th Int’l Conf Intel Transport Syst (ITSC), 2014. Qingdao, China.
[11]	周华, 张巧玉, 母尧尧, 等. 十字路口乘用车与两轮车典型危险场景聚类及推演[J]. 中国安全科学学报, 2020, 30(4):100-107.
	ZHOU Hua, ZHANG Qiaoyu, MU Yaoyao, et al. Clustering and deduction of typical dangerous scenes for passenger cars and two-wheeled vehicles at intersections[J]. *Chin Safe Sci J*, 2020, 30(4):100-107. (in Chinese)
[12]	苏江平, 陈君毅, 王宏雁, 等. 基于中国危险工况的行人交通冲突典型场景提取与分析[J]. 交通与运输(学术版), 2017(1):209-214.
	SU Jiangping, CHEN Junyi, WANG Hongyan, et al. Extraction and analysis of typical scenes of pedestrian traffic conflicts based on dangerous conditions in China[J]. *Traf Transport (Acad Ed)*, 2017(1):209-214. (in Chinses)
[13]	吴斌, 朱西产, 沈剑平, 等. 基于自然驾驶研究的直行追尾危险场景诱导因素分析[J]. 同济大学学报 (自然科学版), 2018, 46(9):1253-1260.
	WU Bin, ZHU Xichan, SHEN Jianping, et al. Analysis of inducing factors in a straight-forward rear-end collision scene based on natural driving research[J]. *J Tongji Univ (Nat Sci Ed)*, 2018, 46(9):1253-1260. (in Chinese)
[14]	李霖, 朱西产, 刘颖, 等. 涉及骑车人的典型交通危险场景[J]. 同济大学学报(自然科学版), 2014, 42(7):1082-1087.
	LI Lin, ZHU Xichan, LIU Ying, et al. Typical traffic danger scenes involving cyclists[J]. *J Tongji Univ (Nat Sci Ed)*, 2014, 42(7):1082-1087. (in Chinese)
[15]	夏澜, 朱西产, 马志雄. 切入型危险工况下自动紧急制动系统的测试场景[C]// 第14届国际汽车交通安全论坛论文集,汽车车身先进设计制造国家重点实验室, 湖南大学, 2017: 142-155.
	XIA Lan, ZHU Xichan, MA Zhixiong. Test scenario of automatic emergency braking system under cut-in dangerous conditions[C]// Infats Proceed 14th Int’l Forum Autom Traff Safe, State Key Laboratory of Advanced Design and Manufacturing for Vehicle Body, Hunan University, 2017: 142-155.
[16]	郭景华, 李克强, 王进, 等. 基于危险场景聚类分析的前车随机运动状态预测研究[J]. 汽车工程, 2020, 42(7):847-853+859.
	GUO Jinghua, LI Keqiang, WANG Jin, et al. Prediction of the random motion state of the preceding vehicle based on cluster analysis of dangerous scenes [J]. *Autom Engi*, 2020, 42(7):847-853+859. (in Chinese)
[17]	刘颖, 贺锦鹏, 刘卫国, 等. 自动紧急制动系统行人测试场景的研究[J]. 汽车技术, 2014(3):35-39.
	LIU Ying, HE Jinpeng, LIU Weiguo, et al. Research on pedestrian test scenarios of automatic emergency braking system[J]. *Autom Tech*, 2014(3):35-39. (in Chinese)
[18]	Bagschik G, Menzel T, maurer M. Ontology based scene creation for the development of automated vehicles [C]// 2018 IEEE Intel Vehic Symp (IV), Jiangsu, 2018: 1813-1820.
[19]	Rocklage E, Kraft H, Karatas A, et al. Automated Scenario Generation for Regression Testing of Autonomous vehicles [C]// IEEE. 2017 IEEE 20th Int’l Conf Intel Transport Syst (ITSC). New York: IEEE, 2017: 476-483.
[20]	Fellner A, Krenn W, Schlick R, et al. Model-based mutation-driven test case generation via heuristic--guided branching search[J]. *ACM Trans Embe Comput Syst*, 2019, 18(1):1-28.
[21]	左国新. 数据分析实验教程[M]. 武汉: 华中师范大学出版社, 2015: 70-77.
	ZUO Guoxin. Experimental Course of Data Analysis [M]. Wuhan: Huazhong Normal Unive Press, 2015: 70-77. (in Chinese)
[22]	何晓群. 多元统计分析(第4版) [M] 北京: 中国人民大学出版社, 2015: 52-57.
	HE Xiaoqun. Multivariate Statistical Analysis (4 Edit) [M]. Beijing: China Renmin Univ Press, 2015: 52-57. (in Chinese)
[23]	百度文库. 公路等级划分标准 [R/OL]. (2018-03-01), https://wenku.baidu.com/view/1cd1b3ed30126edb6f1aff00bed5b9f3f80f72e9.html.
	Baidu Library. Highway classification standards [R/OL]. (2018-03-01), https://wenku.baidu.com/view/1cd1b3ed30126edb6f1aff00bed5b9f3f80f72e9.html. (in Chinese)
[24]	Han I. Scenario establishment and characteristic analysis of intersection collision accidents for advanced driver assistance systems[J]. *Traf Inju Prevent*, 2020, 21(6):1-5.
[25]	Euro-NCAP. TEST PROTOCOL-AEB Car-to-Car systems [EB/OL]. （2020-04-00）, https://www.euroncap.com/en/for-engineers/protocols/safety-assist/.

参数名	内容	取值
天气	晴或多云	0
	阴	1
	雨及其他	2
事故发生时段	日间	0
事故发生时段	夜间	1
参与方1类型	乘用车	0
	两轮车	1
	商用车	2
参与方2类型	乘用车	0
	两轮车	1
	商用车	2
参与方1运动类型	直行	0
	左转	1
	右转	2
参与方2运动类型	直行	0
	左转	1
	右转	2
道路类别	城市道路	0
	乡村道路	1
	公路及其他	2
路口信号灯	直行+转弯信号灯	0
	仅有直行信号灯	1
	无信号灯	2

参数名	内容	取值
天气	晴或多云	0
	阴	1
	雨及其他	2
事故发生时段	日间	0
事故发生时段	夜间	1
参与方1类型	乘用车	0
	两轮车	1
	商用车	2
参与方2类型	乘用车	0
	两轮车	1
	商用车	2
参与方1运动类型	直行	0
	左转	1
	右转	2
参与方2运动类型	直行	0
	左转	1
	右转	2
道路类别	城市道路	0
	乡村道路	1
	公路及其他	2
路口信号灯	直行+转弯信号灯	0
	仅有直行信号灯	1
	无信号灯	2

变量	原取值	转换后取值
晴或多云	0	0	0.5	0.5
阴	1	0.5	0	0.5
雨及其他	2	0.5	0.5	0

变量	原取值	转换后取值
晴或多云	0	0	0.5	0.5
阴	1	0.5	0	0.5
雨及其他	2	0.5	0.5	0

参数	参数	第1类	第2类	第3类	第4类
天气	晴或多云	120	43	22	0
	阴	32	12	3	0
	雨及其他	27	11	5	2
事故发生时段	日间	132	42	28	0
事故发生时段	夜间	47	24	2	2
参与方 1类型	乘用车	142	53	2	0
	两轮车	4	7	1	0
	商用车	33	6	27	2
参与方 2类型	乘用车	41	36	1	0
	两轮车	116	23	29	0
	商用车	22	7	0	2
参与方 1运动类型	直行	169	12	4	1
	左转	3	51	6	0
	右转	7	3	20	1
参与方 2运动类型	直行	86	62	23	1
	左转	79	2	4	0
	右转	14	2	3	1
道路类别	城市道路	51	35	17	0
	乡村道路	22	3	3	2
	公路及其他	106	28	10	0
路口信号灯	直行+ 转弯信号灯	24	18	12	0
	仅有直行信号灯	18	28	7	0
	无信号灯	137	20	11	2
总计		179	66	30	2