基于需求预测与CTM的高速公路交通事件影响范围预测

Prediction of Incident Influence Scope on Expressway Based on Demand Forecasting and CTM

交通事件在高速公路上经常发生,其时间与空间上的不确定性,以及对上游路段和相邻国省道带来的动态衍生影响,使得提前制订具有针对性的预案难度很大。如何快速预测交通事件的影响范围以及交通管控措施的实施效果,成为高速公路应急处置管理的基础。提出基于交通需求预测与元胞传输模型相结合的技术,首先通过卡尔曼滤波算法估计和预测交通需求矩阵,并将上述交通需求加载在元胞传输仿真模型中模拟未来路网的交通运行状态,用于实时、快速预测交通事件发生后未来短时的交通拥堵发展态势。基于该技术,开发了公路网交通运行状态预测系统,通过实际数据的测试,证明该系统在公路网交通事件影响范围预测方面具有良好的精度,并且预测精度随着路网基础交通量的增大而提高。

Incident happened frequently on expressway, the occurrence time and location of which isuncertain. It would cause dynamical derived influence on the upstream section and surrounding network,so that it was impossible to make pre-arranged planning. How to predict the influence area of incidentand the implementaiton effect of the control strategy rapidly became the basis of expressway operationand emergency response. The method which combines demand prediction and cell transmission modelwas proposed for on-line forecasting the short-term propagation of incident on expressway. The trafficdemand predicted by Kalman-Fliter Algorithm was load on CTM Model. Based on this method, evalua.tion and prediction system of state on expressway were developed. Through the test of actual data, it isproved that this system has good accuracy in the prediction of traffic incident influence scope in express.way network, and the accuracy will increase with the raising of network′s basic traffic.