货车碰撞的多层次统计建模与分析

Multilevel Mixed Regression Modeling and Analysis for Truck Collisions

高速公路的快速发展和汽车保有量的日益增加,导致交通事故频发,由于货运车辆的特殊性,使得碰撞所造成的危害程度远远超过其它车辆引发的交通事故。本文针对货车碰撞问题,考虑其主要的影响因素,分析货车的载重量以及车流量的时间分布特征,针对统计数据的阶层结构特点,建立了货车碰撞的多层次混合效应回归模型,结果表明:统计的道路路段越长,路面越宽,交通流量越高,进出口的数量越多,货运量越大,货车发生碰撞的可能性就越大;道路的设计速度越高,安全性要求标准就越高,碰撞的可能性就越小。模型的分析结果可为减少货车的碰撞提供建议。

The development of highways and increasing number of the vehicles has caused the frequent traffic accidents. The impact of this trend is more significant for trucks than smaller vehicles due to the heavy freight weight.Considering the main influencing factors,the distribution characteristics of freight weight and traffic volume were analyzed. In view of the hierarchical structure of statistical data,a multilevel mixed regression model for truck collision is established. The results show that the longer road section,the wider road width and the higher the traffic flow is,the greater possibility of collision will be. Also the collision will enlargement with the increase number of import and export and the freight weight. The higher design speed,the higher security requirement standards,and the smaller possibility of collision will be. The model results can provide suggestions for reducing the truck collisions or the collision degree.