[1]张志坚,江育斌,严利鑫.侧撞事故伤亡决定特征提取和影响因素分析[J].江西师范大学学报(自然科学版),2021,(02):145-152.[doi:10.16357/j.cnki.issn1000-5862.2021.02.06]
　ZHANG Zhijian,JIANG Yubin,YAN Lixin.The Decisive Feature Extraction and Main Influencing Factors Analysis of Side Impact Accidents[J].Journal of Jiangxi Normal University:Natural Science Edition,2021,(02):145-152.[doi:10.16357/j.cnki.issn1000-5862.2021.02.06]
点击复制

侧撞事故伤亡决定特征提取和影响因素分析()
分享到：

《江西师范大学学报》（自然科学版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2021年02期
页码:
145-152
栏目:
信息科学与技术
出版日期:
2021-04-10

文章信息/Info

Title:
The Decisive Feature Extraction and Main Influencing Factors Analysis of Side Impact Accidents
文章编号:
1000-5862(2021)02-0145-08
作者:
张志坚; 江育斌; 严利鑫
华东交通大学交通运输与物流学院,江西 南昌 330013
Author(s):
ZHANG Zhijian; JIANG Yubin; YAN Lixin
School of Transportation and Logistics,East China Jiaotong University,Nanchang Jiangxi 330013,China
关键词:
交通事故;  侧面碰撞;  特征选择;  随机森林模型;  随机参数Logit模型
Keywords:
traffic accident;  side collision;  feature selection;  random forests model;  random parameter logit model
分类号:
TP 311; U 491.31
DOI:
10.16357/j.cnki.issn1000-5862.2021.02.06
文献标志码:
A
摘要:
在交通事故中侧撞事故严重性最强,分析侧撞事故伤亡特征并得出决定性影响因素有助于降低事故风险.该文先使用随机森林特征选择算法对交通事故的影响因素进行降维; 然后,通过随机森林、神经网络、SVM模型3种分类模型来验证降维效果,得出事故影响因素的重要程度; 最后,构建了3种Logit模型来探究事故伤亡程度与影响因素之间的关系,得出各因素对事故严重程度的具体影响.研究结果表明:当提取安全气囊、车辆类型、年龄、防护措施等9种重要程度较大的影响因素作为特征子集时,模型的准确率和有效率达到较高水平.随机参数Logit模型分析的结果表明:在T字路口、驾驶员为男性等情况下事故率较低; 车辆体型越大事故伤亡率越低.
Abstract:
Side collisions are the most serious in traffic accidents.Analyzing the characteristics of injuries and deaths of side collisions and deriving decisive factors can help to reduce the risk of accidents.Firstly,the random forest algorithm is used to reduce the dimensionality of the influencing factors of traffic accidents in this paper.Secondly,the dimensionality reduction effect is verified through three models of classification such as random forest,neural network and SVM,and the importance of the accident influencing factors are obtained.Finally,three Logit models are constructed to explore the relationship between the accident severity and the influencing factors,and the specific impacts of each factor on the severity of the accident are obtained.The results show that the accuracy and efficiency of the model reach a relatively high level when the nine important influencing factors are selected such as airbag,vehicle type,age,and protective measures.The results of the random parameter Logit model show that under the circumstances of lighting at night,T-shaped intersections,the accident rate is lower.The larger the vehicle size is,the lower the accident casualty rate becomes.

参考文献/References:

[1] Yuan Quan,Xu Xuecai,Xu Mingchang,et al.The role of striking and struck vehicles in side crashes between vehicles:Bayesian bivariate probit analysis in China[J].Accident Analysis and Prevention,2020,134:105324.
[2] Celik A K,Oktay E.A multinomial logit analysis of risk factors influencing road traffic injury severities in the Erzurum and Kars Provinces of Turkey[J].Accident Analysis and Prevention,2014,72:66-77.
[3] 沈小燕,魏珊珊,冯煜清.基于机器学习的危险货物道路运输事故影响因素分析[J].交通信息与安全,2020,38(15):113-119,128.
[4] Bareiss M,Gabler H C.Estimating near side crash injury risk in best performing passenger vehicles in the United States[J].Accident Analysis and Prevention,2020,138:105434.
[5] Kelley M E,Talton J W,Weaver A A,et al.Associations between upper extremity injury patterns in side impact motor vehicle collisions with occupant and crash characteristics[J].Accident Analysis and Prevention,2019,122:1-7.
[6] Zhai Ben,Lu Jiantao,Wang Yanli,et al.Real-time prediction of crash risk on freeways under fog conditions[J].International Journal of Transportation Science and Technology,2020,9(4):287-298.
[7] 朱林艳. 基于随机森林的行人交通事故研究[D].武汉:华中科技大学,2018.
[8] 黄兆国,过秀成,贾亮.基于随机森林的雨天车辆跟驰风险行为研究[J].交通信息与安全,2020,38(1):27-34.
[9] 王少华,黄建玲,陈艳艳,等.自行车驾驶员交通行为方式判定研究[J].重庆交通大学学报:自然科学版,2020,39(6):19-24.
[10] 高珍,柯阿香,余荣杰,等.基于随机生存森林的交通事件持续时间预测[J].同济大学学报:自然科学版,2017,45(9):1304-1310.
[11] Hamad K,Al-Ruzouq R,Zeiada W,et al.Predicting incident duration using random forests[J].Transportmetrica A:Transport Science,2020,16(3):1269-1293.
[12] Yu Bo,Chen Yuren,Bao Shan.Quantifying visual road environment to establish a speeding prediction model:an examination using naturalistic driving data[J].Accident Analysis and Prevention,2019,129:289-298.
[13] Ulfarsson G F,Mannering F L.Differences in male and female injury severities in sport-utility vehicle,minivan, pickup and passenger car accidents[J].Accident Analysis and Prevention,2004,36(2):135-147.
[14] 马壮林,张祎祎,杨杨,等.公路隧道交通事故严重程度预测模型研究[J].中国安全科学学报,2015,25(5):75-79.
[15] Liu Pengfei,Fan Wei.Exploring injury severity in head-on crashes using latent class clustering analysis and mixed logit model:a case study of North Carolina[J].Accident Analysis and Prevention,2020,135:105388.
[16] Rezapour M,Wulff S S,Ksaibati K.Examination of the severity of two-lane highway traffic barrier crashes using the mixed logit model[J].Journal of Safety Research,2019,70:223-232.
[17] Hou Qinzhong,Huo Xiaoyan,Leng Junqiang,et al.Examination of driver injury severity in freeway single-vehicle crashes using a mixed logit model with heterogeneity-in-means[J].Physica A:Statistical Mechanics and Its Applications,2019,531:121760.
[18] 陈昭明,徐文远,曲悠扬,等.基于混合Logit模型的高速公路交通事故严重程度分析[J].交通信息与安全,2019,37(3):42-50.
[19] 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.
[20] 李光华,李俊清,张亮,等.一种融合蚁群算法和随机森林的特征选择方法[J].计算机科学,2019,46(11A):212-215.
[21] Vajari M A,Aghabayk K,Sadeghian M,et al.A multinomial logit model of motorcycle crash severity at Australian intersections[J].Journal of Safety Research,2020,73:17-24.
[22] Osman M,Mishra S,Paleti R.Injury severity analysis of commercially-licensed drivers in single-vehicle crashes:accounting for unobserved heterogeneity and age group differences[J].Accident Analysis and Prevention,2018,118:289-300.

备注/Memo

备注/Memo:
收稿日期:2020-10-29
基金项目:国家自然科学基金(51805169)资助项目.
作者简介:张志坚(1978—),男,江西丰城人,副教授,博士,主要从事交通运输、运营与供应链管理研究.E-mail:zzjjxs@126.com
更新日期/Last Update: 2021-04-10