Influence Factors and Coupling Relationship of Traffic Accident
Injury Degree Based on a Data-driven Approach

doi:10.16097/j.cnki.1009-6744.2022.05.012

Abstract

Abstract: In order to accurately identify the relevant factors affecting the traffic accident injury degree of mountainous expressway (TAIDME), a model named random forest naive bayes-coupling degree model (RFNB-CDM) was constructed. Firstly, 1760 pieces of accident data of mountainous expressway in Yunnan Province from 2016 to 2020 were processed. And 18 factors including accident information, road information, accident motor vehicle information, and driver information were studied as initial features. A RF model was used for feature extraction, and the importance ranking of each factor for the severity of traffic accidents (TASME) of mountainous expressway was obtained. Secondly, the new features are input into a NB model to conduct a single factor analysis on the influencing factors of TAIDME. To improve the shortcomings of the original model that cannot accurately describe the relationship between the influencing factors, this paper introduces the coupling degree model to make an example verification analysis. Eight kinds of factors, i.e., rear-end collision, the period from 18:00 to 6:00 of the next day, the number of accident vehicles, downhill, no street lighting at night, freight, large and medium- sized trucks, and straight uniform are more likely to increase TAIDME. The coupling effect of rear- end collision and straight uniform velocity is more likely to lead tomajor accidents. Road surface dryness, roadside metal protection, and central green belt isolation can reduce TAIDME, and when roadside metal protection and central green belt isolation are coupled, the TAIDME can be reduced. The conclusion of this study can provide a theoretical basis and decision- making reference for the prevention of traffic accidents and the reduction of the injury degree of mountain highway accidents.

Key words: traffic engineering, influencing factors, RFNB-CDM, mountain expressway, accident injury degree

摘要： 为准确识别影响山区高速公路交通事故伤害程度(TAIDME)的相关因素，本文构建随机森林朴素贝叶斯-耦合度模型(RFNB-CDM)对其进行研究。首先，处理2016—2020年云南省1760起山区高速公路事故数据，综合考量后，将涉及事故信息、道路信息、肇事机动车辆信息及驾驶人信息等4类18个相关因素作为初始特征进行研究，使用RF模型进行特征提取，并得到各因素对于山区高速公路交通事故严重程度(TASME)的重要性排序；其次，将新特征输入 NB 模型，对TAIDME的影响因素进行单因素分析；为改进原有模型不能对影响因素之间的关系进行准确刻画的缺点，本文引入耦合度模型并进行实例验证分析。结果表明：RFNB模型相较于RF和NB模型，得到的预测结果更加精确，分类性能分别提升5.56%和14.79%。其中，追尾碰撞、18:00-次日 6:00、事故车辆数2辆、下坡段、夜间无路灯照明、货运、大中型货车和直行匀速这8类因素存在时更易加重TAIDME，追尾碰撞和直行匀速这两类因素发生耦合作用时，最易导致重大伤害事故；道路表面干燥、路侧金属防护和中央绿化带隔离这3类因素存在时可降低TAIDME，路侧金属防护和中央绿化带隔离这两类因素发生耦合作用时TAIDME最低。研究结论可为山区高速公路交通事故预防及降低山区高速公路事故发生后的伤害程度提供理论依据与决策参考。

关键词: 交通工程, 影响因素, RFNB-CDM, 山区高速公路, 事故伤害程度

CLC Number:

U491.31

HU Li-wei , LV Yi-fan, ZHAO Xue-ting, XUE Yu, ZHANG Cheng-jie, LEI Guo-qing , LIU Fan. Influence Factors and Coupling Relationship of Traffic Accident Injury Degree Based on a Data-driven Approach[J]. Journal of Transportation Systems Engineering and Information Technology, 2022, 22(5): 117-124.

胡立伟, 吕一帆, 赵雪亭, 薛宇, 张成杰, 雷国庆, 刘凡. 基于数据驱动的交通事故伤害程度影响因素及其耦合关系研究[J]. 交通运输系统工程与信息, 2022, 22(5): 117-124.

Add to citation manager EndNote|Ris|BibTeX

URL: http://www.tseit.org.cn/EN/10.16097/j.cnki.1009-6744.2022.05.012

http://www.tseit.org.cn/EN/Y2022/V22/I5/117

References

[1] BUCSUHÁZY K, MATUCHOVÁ E, ZŮVALA R, et al. Human factors contributing to the road traffic accident occurrence[J]. Transportation Research Procedia, 2020, 45: 555-561.

[2] JONGHAK L, JUNGHYO C, TAEKWAN Y, et al. Traffic accident severity analysis with rain-related factors using structural equation modeling: A case study of Seoul City [J]. Accident Analysis and Prevention, 2018, 112: 1-10.

[3] 戢晓峰, 吴亚欣, 袁华智, 等. 二级公路交通流特征对事故严重程度的影响[J]. 中国公路学报, 2020, 33(2): 135-145. [JI X F, WU Y X, YUAN H Z, et al. Influences of traffic flow characteristics on accident severity on secondary roads[J]. China Journal of Highway and Transport, 2020, 33(2): 135-145.]

[4] 孙轶轩, 邵春福, 赵丹, 等. 交通事故严重程度C5.0决策树预测模型[J]. 长安大学学报(自然科学版), 2014, 34(5): 109-116. [SUN Y X, SHAO C F, ZHAO D, et al. Traffic accident severity prediction model based on C5.0 decision tree[J]. Journal of Chang'an University(Natural Science Edition), 2014, 34(5): 109-116.]

[5] 赵玮, 徐良杰, 冉斌, 等. 基于Adaptive Lasso及RF算法的冰雪天气交通事故分析[J]. 中国安全科学学报, 2017, 27(2): 98-103. [ZHAO W, XU L J, RAN B, et al. Analyzing traffic crash under iced and snow weather condition based on Adaptive Lasso and RF[J]. China Safety Science Journal, 2017, 27(2): 98-103.]

[6] 房锐, 张琪, 胡澄宇, 等. 基于风险矩阵的干线公路弯道路段交通冲突风险评估模型[J]. 交通运输系统工程与信息, 2021, 21(2): 166-172. [FANG R, ZHANG Q, HU C Y, et al. Risk assessment model based on risk matrix for traffic conflict on arterial highway bend section [J]. Journal of Transportation Systems Engineering and Information Technology, 2021, 21(2): 166-172.]

[7] 李英帅, 张旭, 王卫杰, 等. 基于随机森林的电动自行车骑行者事故伤害程度影响因素分析[J]. 交通运输系统工程与信息, 2021, 21(1): 196-200. [LI Y S, ZHANG X, WANG W J, et al. Factors affecting electric bicycle rider injury in accident based on random forest model[J]. Journal of Transportation Systems Engineering and Information Technology, 2021, 21(1): 196-200.]

[8] 张轮, 杨文臣, 刘拓, 等. 基于朴素贝叶斯分类的高速公路交通事件检测[J]. 同济大学学报(自然科学版), 2014, 42(4): 558- 563, 595. [ZHANG L, YANG W C, LIU T, et al. A naive bayesian classifier-based algorithm for freeway traffic incident detection[J]. Journal of Tongji University(Natural science), 2014, 42(4): 558-563, 595.]

[9] ZONG F, XU H G, ZHANG H Y, et al. Prediction for traffic accident severity: Comparing the bayesian network and regression models[J]. Mathematical Problems in Engineering, 2013, 43: 402-411.

[10] 张军, 胡震波, 朱新山. 基于AdaBoost分类器的实时交通事故预测[J]. 计算机应用, 2017, 37(1): 284- 288. [ZHANG J, HU Z B, ZHU X S. Real-time traffic accident prediction based on AdaBoost classifier[J]. Journal of Computer Applications, 2017, 37(1): 284-288.]

[11] ATHANASIOS T. Incorporating real-time traffic and weather data to explore road accident likelihood and severity in urban arterials[J]. Journal of Safety Research, 2017, 61: 9-21.

[12] 张文钧, 蒋良孝, 张欢, 等. 一种双层贝叶斯模型：随机森林朴素贝叶斯[J]. 计算机研究与发展, 2021, 58(9): 2040-2051. [ZHANG W J, JIANG L X, ZHANG H, et al. A two-layer bayes model: Random forest naive bayes[J]. Journal of Computer Research and Development, 2021, 58(9): 2040-2051.]

[13] 中华人民共和国公安部. GA/T 1082-2013, 道路交通事故信息调查[S]. 北京: 中国标准出版社, 2013. [The Ministry of Public Security of the People's Republic of China. GA/T 1082-2013, Information investigation on road traffic accident[S]. Beijing: Standards Press of China, 2013.]

[14] SEYED H A H, SEYED M H H. Traffic accident severity prediction using a novel multi-objective genetic algorithm [J]. Taylor & Francis, 2017, 22: 425-440.

[15] 宋述芳, 何入洋. 基于随机森林的重要性测度指标体系[J]. 国防科技大学学报, 2021, 43(2): 25-32. [SONG S F, HE R Y. Importance measure index system based on random forest[J]. Journal of National University of Defense Technology, 2021, 43(2): 25-32.]