高速公路交通事故持续时间计算方法研究  

Study on Calculation Method of Expressway Accident Duration

作　　者：吕路 李杰 郭忠印[1] 阎莹[3] 高超 Lü Lu;LI Jie;GUO Zhong-yin;YAN Ying;GAO Chao(Key Laboratory of Road and Traffic Engineering of Ministry of Education,Tongji University,Shanghai 201804,China;Shandong Hi-speed Information Group Co.,Ltd.,Jinan Shandong 250100,China;School of Transportation Engineering,Chang'an University,Xi'an Shaanxi 710064,China;Shaanxi Huidetong Municipal Engineering Co.,Ltd.,Xi'an Shaanxi 710086,China)

机构地区：[1]同济大学道路与交通工程教育部重点实验室,上海201804 [2]山东高速信息集团有限公司,山东济南250100 [3]长安大学运输工程学院,陕西西安710064 [4]陕西汇德通市政工程有限公司,陕西西安710086

出　　处：《公路交通科技》2022年第12期155-162,共8页Journal of Highway and Transportation Research and Development

基　　金：国家自然科学基金项目(51978069);山东省交通运输厅科技计划项目(2019B32);陕西省重点研发计划项目(2021KWZ-09)。

摘　　要：交通事故持续时间是衡量交通事故对交通影响程度的一项重要指标,交通事故持续时间计算对制定事故下的交通分流管控方案和降低事故产生的不利影响具有重要作用。统计了贵州省高速公路不同交通事件频数及事件持续时间的分布。由高速公路交通事故数据集、交通流和气象观测资料、高速公路设计文件,建立了交通事故持续时间初始变量集。通过极限梯度提升树进行了特征变量选择和事故持续时间计算模型训练,并进行了特征变量敏感性分析。采用贵州省不同时期的交通事故数据进行了模型时间稳定性检验,采用贵州省和山东省相同时期的交通事故数据进行了模型空间稳定性检验。对所建模型在实际交通环境中的应用方法进行了讨论。结果表明:交通事故是最常发生的一类交通事件,两车碰撞类事故较单方事故持续时间长24%;同回归模型相比,优化后的集成学习算法可提高交通事故持续时间计算结果的准确率;由于不同时期、不同地区交通事故持续时间影响因素可能存在差异,模型时空稳定性分析结果表明,所建模型具有时间稳定性,但不具备空间稳定性;事故持续时间的风险函数表明,最常发生的一类事故持续时间为29 min,根据其特征变量取值,结合变量敏感性分析结果,交管部门可根据事故特征变量,快速进行事故持续时间计算。The duration of traffic accident is an important indicator to measure the impact of traffic accident on traffic.The calculation of the duration of traffic accidents plays an important role in formulating traffic diversion management and control scheme under accident and reducing the impact of accident.The distributions of frequency and duration of different traffic incidents on expressway in Guizhou Province are counted.According to the expressway accident dataset,traffic flow and meteorological observation data and the expressway design document,the initial variable set of traffic accident duration is established.The characteristic variables are selected and the accident duration calculation model training is carried out by using XGBoost,and the sensitivity analysis of the characteristic variable is carried out.The temporal stability of the model is tested by using the traffic accident data of Guizhou Province in different periods,and the spatial stability of the model is tested by using the traffic accident data of Guizhou Province and Shandong Province in the same period.The application method of the established model in the actual traffic environment is discussed.The result shows that(1)Traffic accidents are the most common type of traffic incidents,and the duration of the accident between 2 drivers is 24%longer than the unilateral accident;(2)compared with the regression model,the optimized integrated learning algorithm can improve the accuracy of traffic accident duration calculation;(3)because the influencing factors of accident duration may differ in different periods and regions,the result of the spatial-temporal stability analysis shows that the model has temporal stability but does not have spatial stability;(4)the hazard function of accident duration shows that the duration of the most common accident type is 29 min,according to the value of its characteristic variables and the result of variable sensitivity analysis,the traffic management department can quickly calculate the accident duration based o

关 键 词：交通安全 交通事故 事故持续时间 计算方法 极限梯度提升树 

分 类 号：U491.4[交通运输工程—交通运输规划与管理]