摘要
交通事故的发生具有随机性和偶然性,为尽可能地降低交通事故的伤害程度,根据某高速公路典型事故多发段的交通事故统计资料,以交通事故严重程度为因变量,从时间、道路空间结构和交通运行环境等因素中初步选择12个候选自变量,采用混合逐步选择法分析候选自变量与因变量是否显著相关。采用累积Logistic模型建立交通事故严重程度时空分析模型,并从成比例检验、拟合优度检验和预测准确度检验3个方面对模型进行检验。研究结果表明:事故发生时段、季节因素、发生地点、道路线形、坡度、事故涉及车辆数和日标准交通量与年平均日交通量之比与交通事故严重程度显著相关。
The occurrence of traffic accidents has the randomicity and contingency,this article attempts to make a research on the rules of traffic accidents to minimize the severity of traffic accidents.Firstly,according to the statistical data of a typical accident-prone section,accident severity was selected as the dependent variable,and twelve factors from the three aspects of time,road spatial structure and traffic environment were selected as the candidate independent variables.Then,the combined stepwise method was used to analyze the significant correlation between dependent variable and independent variables,and a temporal-spatial analysis model of traffic accident based on cumulative Logistic model was built.The developed model was tested from three aspects,which are score test for the proportional odds assumption,goodness of fit and predictive accuracy.The results show that seven independent variables,which are accident time,seasonal factors,accident location,road alignment,gradient,the number of vehicles involved in accidents and the ratio of daily traffic to annual average daily traffic,are significantly associated with the dependent variable.
出处
《中国安全科学学报》
CAS
CSCD
北大核心
2011年第9期94-99,共6页
China Safety Science Journal
基金
国家科技支撑计划(2007BAK35B06)
高等学校博士学科点专项科研基金资助(20090009110011)
中央高校基本科研业务费专项资金(CHD2011JC112)
长安大学基础研究支持计划专项资金
作者简介
马壮林（1980-），男，辽宁鞍山人，长安大学讲师，工学博士，研究方向为交通安全。E—mail：zhuanglinma@chd．edu．cn．
