论文部分内容阅读
为了进行山区高速公路追尾事故预测并识别追尾事故突出诱导因素,在对两车追尾事故进行类别划分并确定出典型两车追尾事故的基础上,分析了典型两车追尾事故的事故率与线形指标、车速差、大型车混入率、交通量等单一因素间的相关关系。鉴于单一因素与追尾事故率间的关系不能准确描述追尾事故发生规律的缺陷,建立了线形与交通状态组合条件下的追尾事故次数负二项分布预测模型,并给出了模型变量弹性系数计算方法,用以确定追尾事故的突出诱导因素。研究结果表明:基于线形与交通状态的追尾事故负二项分布预测模型能够对追尾事故进行准确预测,利用弹性系数计算方法确定出车速差、年平均日交通量(AADT)以及竖曲线半径为典型两车追尾事故的突出诱导因素。
In order to make a rear-end accident prediction and identify the rear-end accident highlighting factors in mountainous expressways, based on the classification of two rear-end accidents and the typical two-car rear-end accidents, the accident rates and linear indexes , Speed difference, large-scale car mixed rate, traffic and other single-factor correlation between. In view of the fact that the relationship between single factor and rear-end accident rate can not describe the defect of rear-end accident accurately, the negative binomial distribution forecasting model of rear-end accident number under combination of linear and traffic condition is established and the elastic coefficient of model variable is calculated , To identify the rear-end accidental prominent inducing factor. The results show that the negative binomial distribution forecasting model based on linear and traffic conditions can predict the rear-end accident accurately, and the elastic coefficient calculation method is used to determine the speed difference, annual average daily traffic volume (AADT) and vertical curve radius as typical Two cars rear-end accident prominent highlight factor.