Dynamic Prediction of Traffic Situations and Travel Time on Winter Road Surface

为了弄清冬季路面上的交通流特征，1)利用实际道路上的跟车数据，比较了夏季路面和冬季路面的反应时间、潜在碰撞和冲击波等特征。2)将API模块集成到无迹卡尔曼滤波中，提出了一种实时估计交通状况和交通需求的逆方法。3)建立了基于反馈概念的动态行程时间预测模型，并进行了精度验证。

In order to clarify the characteristics of traffic flow on winter road surfaces,1)some features of reaction time, potential collision, and shockwave were compared with those on summer roads using the car-following data on actual roads. 2) An inverse method of estimating traffic situation and traffic demand on real time was developed by integrating the API module into unscented Kalman filtering. 3) A dynamic travel time prediction model that implemented the above achievements based on feedback concept was developped and justified with the precision.