摘要
随着上海市规划建成区不断向郊区扩大,原有的二级公路很多转变为过境和集散功能兼具的城镇化道路。现收集了2017年上海市33条典型二级公路的历史事故数据和路段基础设施数据,对事故率进行了负二项分布拟合,拟合通过了K-S检验,以事故率为初始因变量,道路基础设施属性为变量,构建二级公路事故预测模型。将事故率进行风险等级划分作为最终因变量,构建事故风险等级预测模型。该模型训练集预测整体准确率为90.3%,测试集预测准确率为89.3%。通过模型结果对上海市二级公路事故致因进行分析,发现:路面宽度、人非隔离形式、机非隔离形式、接入口密度和机非分隔带开口密度对事故风险有显著的影响,应作为设计和管理中的控制对象。
As the planned built-up area of Shanghai continues to expand to the suburbs,many of the original second-class highways have been transformed into urbanized roads with both transit and distribution functions.Accidents of second-class highways in Shanghai are particularly serious.Historical accident data and road infrastructure data for 33 typical second-class highways in Shanghai in 2017 were collected.Negative binomial distribution fitting of annual accidents per kilometer was carried out,which was checked by K-S test.A second-class highway accident prediction model was established with annual accidents per kilometer as the initial dependent variable and road infrastructure attributes as variables.The risk grade of annual accidents per kilometer was used as the final dependent variable to establish the accident risk grade prediction model.The overall prediction accuracy of the model training set was 90.3%and test set was 89.3%.Based on the results of the model,the cause of the second-class highway accident in Shanghai was analyzed.It was found that the width of the road,the isolated form of the sidewalk and non-motorized lane,the isolated form of motorized lane and non-motorized lane,the density of access points and the opening density of the motorized iane and non-motorized lane have a significant impact on the accident risk,and should be taken as important control indicators in design and management.
作者
冯奇
蒋萌露
周小鹏
王俊骅
何卓臻
FENG Qi;JIANG Menglu;ZHOU Xiaopeng;WANG Junhua;HE Zhuozhen
出处
《上海公路》
2020年第3期88-92,M0006,共6页
Shanghai Highways
关键词
负二项分布
二级公路
风险等级
事故致因
交通工程
Negative binomial distribution
second-class highway
risk grade
accident cause
traffic engineering
