考虑非均衡性的城市自行车事故骑行者伤害程度影响因素及异质性分析

doi:10.3976/j.issn.1002-4026.2025034

摘要/Abstract

摘要： 为探究城市自行车事故骑行者伤害程度的影响因素，同时降低数据异质性和非均衡性对因素量化的影响。基于CRSS数据库的3 895起自行车事故，提出了一种融合重采样、潜在类别分析（LCA）和Bayes网络（BN）的方法。首先，采用LCA将事故数据重新划分为若干组具有组内同质性和组间异质性的子事故群，减少数据异质性的影响；其次，采用随机过采样（ROS）、合成少数类过采样技术（SMOTE）和自适应合成过采样算法（ADASYN）对各事故群重采样，减少数据非均衡性的影响；最后，基于各类重采样后的事故群，分别搭配2种BN结构学习算法和1种参数学习算法，并依据AUC值评选每类事故群的最优BN模型，实现骑行者伤害程度影响因素的定量分析和异质性分析。研究结果表明：当整体事故数据被划分为3类同质子数据群时，LCA模型的Entropy值较优，达0.943。其中C1事故群、C2事故群、C3事故群和OD事故群分别被挖掘出10、13、9和12个影响骑行者伤害程度的关键因素；将LCA和重采样引入BN，能显著提升BN模型的G-mean值、AUC值和风险因素挖掘能力；时间段、骑行者性别、骑行者年龄和天气状况等因素在不同事故群中存在明显的异质性。

关键词: 交通安全, 自行车事故, 伤害程度, 潜在类别分析, Bayes网络

Abstract: To explore the factors influencing the injury severity of cyclists in urban bicycle accidents and mitigate the impact of data heterogeneity and imbalance on the quantification of these factors, this study proposes a method integrating resampling, latent class analysis (LCA), and Bayesian networks (BNs) based on 3 895 bicycle accidents from the CRSS database. First, LCA was used to reclassify accident data into several sub-accident clusters with intra-cluster homogeneity and inter-cluster heterogeneity to reduce the impact of data heterogeneity. Second, random over-sampling (ROS), synthetic minority oversampling technique, and adaptive synthetic sampling approach were used to resample each accident cluster to reduce the impact of data imbalance. Finally, based on various resampled accident clusters, two BN structure learning algorithms and one parameter learning algorithm were applied and the optimal BN model for each accident cluster was selected based onAUC values to enable quantitative and heterogeneity analyses of factors influencing the injury severity of cyclists. Results show that when the overall accident data were divided into three homogeneous sub-clusters, the LCA model achieved an increased entropy value of 0.943. For the C1, C2, C3, and OD accident clusters, 10, 13, 9, and 12 key factors influencing the injury severity of cyclists were identified, respectively. The introduction of LCA and resampling into the BN considerably improved the BN model’sG-meanvalue,AUC value, and risk factor identification capabilities. Factors such as time period, cyclist’s gender, cyclist’s age, and weather conditions showed substantial heterogeneity across different accident clusters.

Key words: traffic safety, bicycle accidents, injury severity, latent class analysis, Bayesian networks

中图分类号:

U491.31

王朝健, 徐小金, 冯斌, 余松霖, 张卫东. 考虑非均衡性的城市自行车事故骑行者伤害程度影响因素及异质性分析[J]. 山东科学, 0, (): 1-.

WANG Chaojian, XU Xiaojin, FENG Bin, YU Songlin, ZHANG Weidong. Analysis of factors influencing cyclist injury severity and heterogeneity analysis in urban bicycle accidents considering data imbalance[J]. Shandong Science, 0, (): 1-.

参考文献

[1] 王丙雨，邹俊，韩勇，等. 车辆和自行车碰撞事故中骑车人下肢损伤风险研究[J]. 振动与冲击, 2023, 42(11): 324-330. DOI: 10.13465/j.cnki.jvs.2023.11.038.

[2] 魏晋，安实，张炎棠. 考虑建成环境交互影响的共享单车需求预测[J]. 科学技术与工程, 2023, 23(26): 11424-11430.

[3] 赵琳娜，贾兴无，戴帅，等.中国城市道路交通安全特点解析[J]. 城市交通, 2018, 16(03): 9-14+20. DOI: 10.13813/j.cn11-5141/u.2018.0302.

[4] 黎健侃，李泽炜，华汶雯，等. 城市道路交通事故统计分析[J]. 科技创新与应用, 2021, 11(21): 74-76. DOI: 10.19981/j.cn23-1581/g3.2021.21.023.

[5] ALNAWMASI N, MANNERING F. An analysis of day and night bicyclist injury severities in vehicle/bicycle crashes: A comparison of unconstrained and partially constrained temporal modeling approaches[J]. Analytic methods in accident research, 2023, 40: 100301. DOI: 10.1016/j.amar.2023.100301.

[6] WU J, RASOULI S, ZHAO J, et al. Large truck fatal crash severity segmentation and analysis incorporating all parties involved: A Bayesian network approach[J]. Travel Behaviour and Society, 2023, 30: 135-147. DOI: 10.1016/j.tbs.2022.09.003.

[7] LUAN S, LI M, LI X, et al. Effects of built environment on bicycle wrong Way riding behavior: A data-driven approach[J]. Accident Analysis & Prevention, 2020, 144: 105613. DOI: 10.1016/j.aap.2020.105613.

[8] DASH I, ABKOWITZ M, PHILIP C. Factors impacting bike crash severity in urban areas[J]. Journal of safety research, 2022, 83: 128-138. DOI: 10.1016/j.jsr.2022.08.010.

[9] YANG Z, YANG Z, SMITH J, et al. Risk analysis of bicycle accidents: A Bayesian approach[J]. Reliability Engineering & System Safety, 2021, 209: 107460. DOI: 10.1016/j.ress.2021.107460.

[10] 杨园园，鲁统宇，崔俊，等. 考虑错分代价的ADASVM-CSLINEX模型及应用[J]. 计算机工程与应用, 2024, 60(03): 348-356. DOI: 10.3778/j.issn.1002-8331.2210-0379.

[11]谭鼎. 基于集成学习的慢行交通事故严重程度预测及致因分析[D]. 兰州：兰州交通大学, 2024. DOI: 10.27205/d.cnki.gltec.2024.001534.

[12] YAHAYA M, GUO R, FAN W, et al. Bayesian networks for imbalance data to investigate the contributing factors to fatal injury crashes on the Ghanaian highways[J]. Accident Analysis & Prevention, 2021, 150: 105936. DOI: 10.1016/j.aap.2020.105936.

[13] 潘义勇,李烁.建成环境对交叉口行人事故严重程度异质性影响[J].重庆交通大学学报(自然科学版),2024,43(06):87-93+117. DOI: 10.3969/j.issn.1674-0696.2024.06.12.

[14] HOSSEINI S H, DAVOODI S R, BEHNOOD A. Bicyclists injury severities: An empirical assessment of temporal stability[J]. Accident Analysis & Prevention, 2022, 168: 106616. DOI: 10.1016/j.aap.2022.106616.

[15] LIN Z, FAN W D. Exploring bicyclist injury severity in bicycle-vehicle crashes using latent class clustering analysis and partial proportional odds models[J]. Journal of safety research, 2021, 76: 101-117. DOI: 10.1016/j.jsr.2020.11.012.

[16] 马硕.考虑时空效应的自行车事故特征及影响因素研究[D].北京:北京建筑大学, 2024. DOI: 10.26943/d.cnki.gbjzc.2024.000816.

[17] SONG Y, CHITTURI M V, NOYCE D A. Intersection two-vehicle crash scenario specification for automated vehicle safety evaluation using sequence analysis and Bayesian networks[J]. Accident Analysis & Prevention, 2022, 176: 106814. DOI: 10.1016/j.aap.2022.106814.

[18] 邓佳.人—车交通事故严重程度Probit预测模型构建及实证研究[D].西安：长安大学,2019.

[19] 晏钰棚. 自行车事故伤害严重程度影响因素相对重要性和时空异质性研究[D]. 成都：西南交通大学, 2021. DOI: 10.27414/d.cnki.gxnju.2021.001225.

[20] 布和.道路交通事故的成因分析及预防研究[J].武汉公安干部学院学报, 2019, 33(02): 16-20.

[21] 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(05): 24-28.

[22] 孙少峰. 基于集成学习的新能源车辆与内燃机车辆交通事故严重程度影响因素对比研究[D]. 西安：长安大学, 2021. DOI: 10.26976/d.cnki.gchau.2021.001818.

[23] 翟洪军，陈启光，申春悌，等. 基于潜在类别分析对不同年龄组患者新冠肺炎病因病机证候研究[J].世界科学技术-中医药现代化, 2021, 23(03): 866-873. DOI: 10.11842/wst.20200622001.

[24] 焦朋朋，李汝鉴，王健宇，等. 考虑潜在类别的老年行人交通事故严重程度致因分析[J].交通运输系统工程与信息, 2022, 22(05): 328-336. DOI: 10.16097/j.cnki.1009-6744.2022.05.034.

[25] 李昂, 韩萌, 穆栋梁, 等. 多类不平衡数据分类方法综述[J]. 计算机应用研究, 2022, 39(12): 3534-3545. DOI: 10.19734/j.issn.1001-3695.2022.03.0198.

[26] ABD ELRAHMAN S M, ABRAHAM A. A review of class imbalance problem[J]. Journal of Network and Innovative Computing, 2013, 1: 9-9.

[27] DELEN D, TOMAK L, TOPUZ K, et al. Investigating injury severity risk factors in automobile crashes with predictive analytics and sensitivity analysis methods[J]. Journal of Transport & Health, 2017, 4: 118-131.

[28] HE H, BAI Y, GARCIA E A, et al. ADASYN: Adaptive synthetic sampling approach for imbalanced learning[C]//2008 IEEE international joint conference on neural networks (IEEE world congress on computational intelligence). Ieee, 2008: 1322-1328.

[29] CHAWLA N V, BOWYER K W, HALL L O, et al. SMOTE: synthetic minority over-sampling technique[J]. Journal of artificial intelligence research, 2002, 16: 321-357. DOI: 10.1613/jair.953.

[30] HECKERMAN D, GEIGER D, CHICKERING D M. Learning Bayesian networks: The combination of knowledge and statistical data[J]. Machine learning, 1995, 20: 197-243. DOI: 10.1007/BF00994016.

[31] SIVASANKARAN S K, BALASUBRAMANIAN V. Exploring the severity of bicycle–vehicle crashes using latent class clustering approach in India[J]. Journal of safety research, 2020, 72: 127-138. DOI: 10.1016/j.jsr.2019.12.012.

[32] SUN Z, XING Y, WANG J, et al. Exploring injury severity of bicycle-motor vehicle crashes: A two-stage approach integrating latent class analysis and random parameter logit model[J]. Journal of Transportation Safety & Security, 2022, 14(11): 1838-1864. DOI: 10.1080/19439962.2021.1971814.

[33] SAMEREI S A, AGHABAYK K, SHIWAKOTI N, et al. Using latent class clustering and binary logistic regression to model Australian cyclist injury severity in motor vehicle–bicycle crashes[J]. Journal of Safety Research, 2021, 79: 246-256. DOI: 10.1016/j.jsr.2021.09.005.

[34]孙晴.考虑时间不稳定性的货车-小汽车事故严重程度影响因素分析[D].西安：长安大学, 2021. DOI: 10.26976/d.cnki.gchau.2021.000866.

[35] MYHRMANN M S, JANSTRUP K H, MØLLER M, et al. Factors influencing the injury severity of single-bicycle crashes[J]. Accident Analysis & Prevention, 2021, 149: 105875. DOI: 10.1016/j.aap.2020.105875.

[36] 王精滢.考虑空间异质性的机非交通事故严重程度分析[D]. 成都：西南交通大学, 2020. DOI: 10.27414/d.cnki.gxnju.2020.000850.

[37] 丁晶玉.考虑关联因素异质性及非线性的骑行者碰撞事故受伤严重程度研究[D]. 大连：大连交通大学,2024. DOI: 10.26990/d.cnki.gsltc.2024.000296.