Research on severity of mountain highway accidents based on random parameter logit model

doi:10.16265/j.cnki.issn1003-3033.2024.04.0448

Abstract

Abstract:

In order to investigate the influence of flat curve, longitudinal curve, and flat-long combination of line shapes on the severity of traffic accidents on mountain highways, firstly, the line shape data of a 263 km section of a mountain highway and the 1 870 accidents occurring on the section in five years were statistically analyzed and classified. Then, according to the structural form of the research data, the stochastic parametric logit model was established, and then the model was utilized to study the effect of the combination of line shapes on the severity of accidents. The results show that: the factors of ratio of curve (ROC), weighted curvature (CW), and delta slope (ΔS₂) are associated with an increase in property damage-only accidents, whereas the combination of slope lengths (CSL) is associated with a decrease in property damage-only accidents; the weighted curvature (CW=0.6-0.7) and continuous uphill slope are associated with an increase in accidents with minor injuries, and the factors of weighted curvature (CW=0.3-0.4), delta slope (ΔS₁), and ratio of curve (ROC) were associated with a decrease in accidents with minor injuries; the variation rate of curvature (VRC), contiguous radius of front and behind (CRF, CRB), and combination of slope and radius (CSR) were associated with an increase in accidents with serious injuries or fatalities, and the ratio of curve (ROC) was associated with a decrease in accidents with serious injuries or fatalities; and the combination of linear shapes had a significant effect on the severity of accidents.

Key words: random parameter logit model, mountain highway, severity of accident, horizontal and vertical combined alignments, property loss accident

CLC Number:

X951交通运输安全

ZHANG Wei, GAO Hanhao, TANG Xinqi, XU Kunmiao. Research on severity of mountain highway accidents based on random parameter logit model[J]. China Safety Science Journal, 2024, 34(4): 101-110.

Figures/Tables 7

Fig.1

Tab.1

Classification and numerical value of independent variables and dependent variables

变量类别	变量名称	虚拟变量符号	变量定义	频数	占比/ %	变量类别	变量名称	虚拟变量符号	变量定义	频数	占比/ %
因变量	事故严重程度	Y₁	仅财产损失事故	1 580	86.63	纵面线形组合	ΔS₁/ %	X₆₃	1.5~2	373	20.43
		Y₂	轻伤事故	125	6.84			X₆₄	2~3	255	13.96
		Y₃	重伤或死亡事故	119	6.52			X₆₅	3~4	135	7.38
平面线形组合	ROC	X₁₁	0~0.4	177	9.68			X₆₆	>4	138	7.59
		X₁₂	0.4~0.5	160	8.77		ΔS₂/ %	参照	<0.5	117	6.42
		X₁₃	0.5~0.6	267	14.65			X₇₁	0.5~1.0	373	20.43
		参照	0.6~0.7	198	10.86			X₇₂	1.0~1.5	440	24.12
		X₁₄	0.7~0.8	313	17.17			X₇₃	1.5~2.0	421	23.10
		X₁₅	0.8~0.9	195	10.70			X₇₄	2~3	251	13.74
		X₁₆	0.9~1.0	514	28.18			X₇₅	3~4	138	7.54
	CW	参照	0~0.2	295	16.20			X₇₆	>4	85	4.65
		X₂₁	0.2~0.3	238	13.05		坡度变化	X₁₂₁	连续上坡	268	14.71
		X₂₂	0.3~0.4	122	6.68			X₁₂₂	连续下坡	899	49.30
		X₂₃	0.4~0.5	236	12.94			参照	先上坡后下坡再上坡	62	3.42
		X₂₄	0.5~0.6	234	12.83			X₁₂₃	先下坡后上坡再下坡	70	3.85
		X₂₅	0.6~0.7	219	12.03			X₁₂₄	先上坡后连续下坡	204	11.18
		X₂₆	0.7~0.9	174	9.52			X₁₂₅	先下坡后连续上坡	75	4.12
		X₂₇	>0.9	305	16.74			X₁₂₆	先连续上坡后下坡	173	9.47
	VRC/ ((°)· km^-1)	参照	0 ~30	175	9.57			X₁₂₇	先连续下坡后上坡	72	3.96
		X₃₁	30~40	253	13.85	平纵线形组合	CSL/ (%· km^-1)	X₄₁	<1	395	21.66
		X₃₂	40 ~50	322	17.65			参照	1 ~1.5	225	12.35
		X₃₃	50 ~60	229	12.57			X₄₂	1.5 ~2	264	14.49
		X₃₄	60 ~70	179	9.79			X₄₃	2 ~2.5	215	11.76
		X₃₅	70 ~80	98	5.40			X₄₄	2.5 ~3	195	10.70
		X₃₆	80 ~100	110	6.04			X₄₅	3 ~4	226	12.41
		X₃₇	>100	458	25.13			X₄₆	>4	303	16.63
	CRF	X₈₁	0~0.3	248	13.58		CSR/ (%· km^-1)	参照	<1	231	12.67
		X₈₂	0.3~0.6	370	20.27			X₅₁	1 ~2	254	13.90
		X₈₃	0.6~0.9	588	32.25			X₅₂	2 ~3	343	18.82
		参照	0.9~1.2	536	29.41			X₅₃	3 ~4	244	13.37
		X₈₄	>1.2	82	4.49			X₅₄	4 ~5	50	2.73
	CRB	X₉₁	0~0.3	263	14.44			X₅₅	5 ~6	149	8.18
		X₉₂	0.3~0.6	340	18.66			X₅₆	6 ~10	186	10.21
		X₉₃	0.6~0.9	481	26.36			X₅₇	>10	367	20.11
		参照	0.9~1.2	624	34.22		VHR	X₁₀₁	0~10	234	12.83
		X₉₄	>1.2	115	6.31			X₁₀₂	10~15	252	13.80
	CT	X₁₁₁	1	646	35.40			X₁₀₃	15~20	186	10.21
	CT	参照	0	1 178	64.60			X₁₀₄	20~30	245	13.42
纵面线形组合	ΔS₁/%	参照	>0.5	127	6.95			X₁₀₅	30~40	193	10.59
		X₆₁	0.5~1.0	400	21.93			X₁₀₆	40~50	251	13.74
		X₆₂	1.0~1.5	397	21.76			参照	>50	463	25.40

Tab.1

Tab.2

Fig.2

Fig.3

Fig.4

Tab.3

References 26

[1]	JTG B01—2014,公路工程技术标准[S].
	JTG B01-2014,Technical standard of highway engineering[S].
[2]	JTG D20—2017,公路路线设计规范[S].
	JTG D20-2017,Design specification for highway alignment[S].
[3]	赵华祥, 杜飞翔, 付开华, 等. 山区公路小半径弯道路段事故严重度影响因素及其异质性比较分析[J]. 交通信息与安全, 2022, 40(3):42-50.
	ZHAO Huaxiang, DU feixiang, FU Kaihua, et al. A comparative analysis of heterogeneous effects of various factors on accident severity at sharp curvesections of mountainous highway[J]. Journal of Transport Information and Safety, 2022, 40(3):42-50.
[4]	WU Qiong, BOGUS S M, CHEN Feng, et al. Mixed logit model-based driver injury severity investigations in single- and multi-vehicle crashes on rural two-lane highways[J]. Accident Analysis & Prevention, 2014, 72: 105-115.
[5]	FOUNTAS G, SARWAR M T, ANASTASOPOULOS P C, et al. Analysis of stationary and dynamic factors affecting highway accident occurrence: a dynamic correlated grouped random parameters binary logit approach[J]. Accident Analysis & Prevention, 2018, 113: 330-340.
[6]	ZHANG Guangnan, LI Yanyan, KING M J, et al. Overloading among crash-involved vehicles in China: identification of factors associated with overloading and crash severity[J]. Traffic Injury Prevention, 2019, 25: 36-46.
[7]	CHANG Fangrong, LI Maosheng, XU Pengpeng, et al. Injury severity of motorcycle riders involved in traffic crashes in hunan, china: a mixed ordered logit approach[J]. International Journal of Environmental Research and Public Health, 2016, 13: 714-728.
[8]	郭延永, 刘攀, 吴瑶, 等. 考虑异质性的贝叶斯交通冲突模型[J]. 中国公路学报, 2018, 31(4): 296-303.
	GUO Yanyong, LIU Pan, WU Yao, et al. Bayesian traffic conflict model accounting for heterogeneity[J]. China Journal of Highway and Transport, 2018, 31(4): 296-303.
[9]	HALEEM K, GAN A. Effect of driver's age and side of impact on crash severity along urban freeways: a mixed Logit approach[J]. Journal of Safety Research, 2013, 46: 67-76. doi: 10.1016/j.jsr.2013.04.002 pmid: 23932687
[10]	SHAHEED M S, GKRITZA K, ZHANG Wei, et al. A mixed logit analysis of two-vehicle crash severities involving a motorcycle[J]. Accident Analysis & Prevention, 2013, 61: 119-128.
[11]	MILTON J, SHANKAR V, MANNERING F. Highway accident severities and the mixed logit model: an exploratory empirical analysis[J]. Accident Analysis and Prevention, 2008, 40(1): 260-266.
[12]	ELURU N, BHAT C, HENSHER D. A mixed generalized ordered response model for examining pedestrian and bicyclist injury severity level in traffic crashes[J]. Accident Analysis and Prevention, 2008, 40 (3):1033-1054.
[13]	SHAHEED M, GKRITZA K. A latent class analysis of single-vehicle motorcycle crash severity outcomes[J]. Analytic Methods in Accident Research, 2014, 2: 30-38.
[14]	刘忠, 田莎, 陈青, 等. 随机系数LOGIT模型的最新发展及其应用[J]. 经济学动态, 2012(12):125-130.
	LIU Zhong, TIAN Sha, CHEN Qing, et al. Recent developments and applications of logit models with random coefficients[J]. Economic Perspectives, 2012 (12): 125-130.
[15]	秦萍, 陈颖翱, 徐晋涛, 等. 北京居民出行行为分析:时间价值和交通需求弹性估算[J]. 经济地理, 2014, 34(11): 17-22.
	QIN Ping, CHEN Ying'ao, XU Jintao, et al. Travel behavior analysis for the residents in Beijing: value of time and travel demand elasticity estimates[J]. Economic Geography, 2014, 34(11): 17-22.
[16]	宋栋栋, 杨小宝, 祖兴水, 等. 基于均值异质性随机参数Logit模型的城市道路事故驾驶员受伤严重程度研究[J]. 交通运输系统工程与信息, 2021, 21(3): 214-220.
	SONG Dongdong, YANG Xiaobao, ZU Xingshui, et al. Examination of driver injury severity in urban crashes: a random parameters logit model with heterogeneity in means approach[J]. Journal of Transportation Systems Engineering and Information Technology, 2021, 21(3): 214-220.
[17]	HASSAN Y, EASA S M. Effect of vertical alignment on driver perception of horizontal curves[J]. Journal of Transportation Engineering, 2003, 129(4): 399-407.
[18]	杨佩佩, 熊坚, 何扬帆. 公路小半径曲线路段驾驶人焦虑水平模型[J]. 中国安全科学学报, 2023, 33(1):227-232. doi: 10.16265/j.cnki.issn1003-3033.2023.01.0508
	YANG Peipei, XIONG Jian, HE Yangfan. Anxiety level model of drivers on sharp circular curve road[J]. China Safety Science Journal, 2023, 33(1):227-232. doi: 10.16265/j.cnki.issn1003-3033.2023.01.0508
[19]	张韡, 魏朗, 彭绍勇. 基于能量守恒原理的公路线形安全性评价[J]. 交通运输工程学报, 2009, 9(4): 85-91.
	ZHANG Wei, WEI Lang, PENG Shaoyong. Safety evaluation of road alignment based on energy conservation law[J]. Journal of Traffic and Transportation Engineering, 2009, 9(4): 85-91.
[20]	孟祥海, 由冰玉, 邱志雄, 等. 基于事故修正系数的山区高速公路事故预测模型[J]. 中国安全科学学报, 2023, 33(1):10-17. doi: 10.16265/j.cnki.issn1003-3033.2023.01.0472
	MENG Xianghai, YOU Bingyu, QIU Zhixiong, et al. Accident prediction model of mountainous freeway based on crash modification[J]. China Safety Science Journal, 2023, 33(1):10-17. doi: 10.16265/j.cnki.issn1003-3033.2023.01.0472
[21]	雷桂荣, 刘峰, 谢华昌, 等. 高速公路交通安全与道路线形和主要交通流特性的关系[J]. 公路与汽运, 2018(5): 41-46.
	LEI Guirong, LIU Feng, XIE Huachang, et al. Relationship between highway traffic safety and road alignment and main traffic flow characteristics[J]. Highway and Automobile Transportation, 2018(5): 41-46.
[22]	赵跃峰, 张生瑞, 马壮林. 基于部分优势比的公路隧道交通事故严重程度分析模型[J]. 中国公路学报, 2018, 31(9): 159-166.
	ZHAO Yuefeng, ZHANG Shengrui, MA Zhuanglin. Analysis of traffic accident severity on highway tunnels using the partial proportion odds model[J]. China Journal of Highway and Transport, 2018, 31(9): 159-166.
[23]	申昕, 沈金星, 郑长江, 等. 基于Multinomial Logit模型的美国北卡罗莱纳州慢行交通事故严重程度分析[J]. 交通与运输, 2021, 37(5): 24-28.
	SHEN Xin, SHEN Jinxing, ZHENG Changjiang, et al. Severity analysis of slow traffic accidents in north carolina based on multinomial logit model[J]. Traffic & Transportation, 2021, 37(5): 24-28.
[24]	姚荣涵, 龙梦, 张文松, 等. 基于潜在类别-Logit模型的共享自动驾驶汽车使用意向[J]. 交通信息与安全, 2022, 40(2): 135-144.
	YAO Ronghan, LONG Meng, ZHANG Wensong, et al. User preferences for shared autonomous vehicles based on latent-class Logit models[J]. Journal of Transport Information and Safety, 2022, 40(2): 135-144.
[25]	温惠英, 汤左淦. 道路交叉口单车事故严重程度影响因素分析[J]. 公路工程, 2019, 44(2): 55-61, 102.
	WEN Huiying, TANG Zuogan. Analysis of crash injury severity in single-vehicle crashes occurring at intersections[J]. Highway Engineering, 2019, 44(2): 55-61, 102.
[26]	陈荔, 李聪颖, 詹立, 等. 城市道路交通事故形态影响因素分析与预测[J]. 长安大学学报:自然科学版, 2022, 42(4): 98-107.
	CHEN Li, LI Congying, ZHAN Li, et al. Influencing factors analysis and prediction of urban road traffic accident patterns[J]. Journal of Chang'an University:Natural Science Edition, 2022, 42(4): 98-107.

事故严重程度	变量	变量定义	参数估计	标准差
仅财产损失事故Y₁	X₁₃	ROC=0.5~0.6	1.314^***	0.198
	X₂₃	CW=0.4~0.5	1.117^***	0.184
	X₄₆/(%·km^-1)	CSL>4	-1.182^***	0.223
	X₇₆/%	ΔS₂>4	0.913^**	0.341
	X₁₂₃	先下坡后上坡再下坡	0.786^**	0.366
轻伤事故Y₂	X₁₂	ROC=0.4~0.5	-1.911^***	0.405
	X₂₂(均值)	CW=0.3~0.4	-2.317^***	1.098
	X₂₂(标准差)	CW=0.3~0.4	3.185^***	1.155
	X₂₅	CW=0.6~0.7	1.416^***	0.303
	X₄₅/(%·km^-1)	CSL=3~4	-1.608^***	0.296
	X₆₄/%	ΔS₁=2~3	-1.968^***	0.274
	X₆₅/%	ΔS₁=3~4	-0.727^**	0.294
	X₇₅/%	ΔS₂=3~4	-1.493^***	0.264
	X₁₂₁	连续上坡	0.843^***	0.227
重伤或死亡事故Y₃	ASC_S3	常数项	-1.863^***	0.132
	X₁₄	ROC=0.7~0.8	-0.947^***	0.362
	X₃₆/((°)· km^-1)	VRC=80~100	0.648^*	0.359
	X₅₇/(%·km^-1)	CSR>10	0.476^*	0.275
	X₈₁	CRF<0.3	0.346^***	0.350
	X₉₂	CRB=0.3~0.6	0.538^**	0.229

拟合优度指标	随机参数logit模型
LL(0)	-2 048.91
LL(β)	-1 067.78
χ²	1 962.26
AIC	2 175.56
BIC	2 286.23
Mcfadden Pseudo R²	0.48