城市快速路互通立交分流区交通冲突预测模型

The Predicting Model of Traffic Conflicts in Diverging Segments of Expressway Interchange

为研究互通立交分流区交通冲突特性,确保车辆的安全运行,提升通行效率,分别选取TTC和PET模型作为判定追尾冲突和换道冲突的标准,通过对交通流和交通冲突数据的分析,拟合出预测交通冲突发生概率的二分类Logistic回归模型。研究结果表明:分流区交通冲突发生数不仅随着交通流率的增加而上升,还受到车辆驶入分流区时所处的位置、换道频次和速度的影响,此外,冲突与非冲突车辆的车头时距之间也存在着显著的差异性。

In order to study the traffic conflicts characteristics of interchanges diverging segment,and to ensure the safe operation of vehicles and improve the traffic efficiency,the TTC and PET models were selected respectively as criterions to judge rear-end collision and lane changing collision,traffic flow and traffic conflict data were analyzed by various statistical means,and a binary Logistic regression model was also constructed to predict the probability of traffic conflict.The results show that the occurrence of traffic conflicts in the diverging segment not only increases with the increase of traffic flow rate,but also is affected by the position,lane change frequency and speed of vehicles when entering the diverging segment.In addition,there is a significant difference between the time headway of conflicted and non-conflicted vehicles.