高速公路交通事故致因网络构建与评价研究  

Research on construction and evaluation of freeway traffic accident cause network

作　　者：唐智慧[1,2] 黄镜入 TANG Zhihui;HUANG Jingru(School of Transportation&Logistics,Southwest Jiaotong University,Chengdu 610031,China;National Engineering Laboratory of Integrated Transportation Big Data Application Technology,Southwest Jiaotong University,Chengdu 610031,China)

机构地区：[1]西南交通大学交通运输与物流学院,成都610031 [2]西南交通大学综合交通大数据应用技术国家工程实验室,成都610031

出　　处：《安全与环境学报》2023年第12期4255-4263,共9页Journal of Safety and Environment

摘　　要：为更加准确地辨识出高速公路交通事故关键致因,基于复杂网络提出了综合属性因素(如性别、年龄、时间等)和风险状态因素(如疲劳驾驶、超速等)的高速公路交通事故致因网络构建方法;针对该异质网络特点,提出了网络风险传播强度、风险节点主动/被动影响力、属性节点紧密度、风险节点中介强度等新的网络拓扑评价指标。以高速公路重特大交通事故为例,依托收集的163份国内高速公路重特大交通事故调查报告对其进行网络构建与致因评价。结果表明:基于复杂网络提出的网络构建方法与拓扑评价指标适用于高速公路交通事故致因的评价;未保持安全距离、临危措施不当和低速行驶等风险中介强度靠前的致因为高速公路重特大交通事故的关键致因。To identify the key causative causes of freeway traffic accidents more accurately,this paper presents the construction method of a freeway traffic accident causative network based on a complex network,integrating attribute factors (such as gender,age,time,etc.) and risk state factors (such as fatigue driving,speeding,etc.).According to the characteristics of the heterogeneous network,and based on the existing causative network topology evaluation indexes,this paper proposed new network topology evaluation indicators,such as network risk propagation strength,risk node active influence,risk node passive influence,attribute node tightness,and risk node mediation strength.Taking the serious traffic accidents on freeways as an example,a total of 51 risk state factors in terms of driver,vehicle,road,environment and management and 5attribute factors such as gender,age,time,vehicle type and location were extracted from 163 investigation reports on serious freeway accidents in China from 2014 to 2022 collected from the official website of the emergency management department.Then,this paper constructed the cause network of serious traffic accidents on freeways according to the proposed network construction method.Network risk propagation strength was used to quantify the risk of the network spread at the overall level.Risk node active/passive influence on risk effective partition was used to divide the node type (output type,transmission type,and input type).Attribute node tightness was used to quantify the degree of risk relationship between attribute nodes.Risk node mediation strength was used to realize the importance ranking of risk nodes.Finally,this paper identified the key causes of serious traffic accidents on freeways.The results show that the network construction method and topology evaluation index based on the complex network is suitable for the evaluation of freeway traffic accident causes.Failure to maintain a safe distance,inappropriate emergency measures,and low-speed driving are the key causes of serious t

关 键 词：安全工程 高速公路 致因评价 复杂网络 重特大交通事故 

分 类 号：X951[环境科学与工程—安全科学]