摘要
智能运输系统(ITS)是当前解决交通问题的最佳途径之一,而动态交通信息的发布与预测是ITS的关键所在.然而由于在各种外部因素下,道路交通状态呈现出随机变化的特性,使得动态交通信息的发布与预测成为一大难点.结合某城市的实际情况,本文建立了在定位误差、采样间隔、车辆类型、道路类型等因素影响下的行程车速估计模型,并用实测数据对模型进行精度验证.结合出租车车辆的特殊信息,改进了低采样频率浮动车技术的估计算法,建立了空车数据处理模型;针对城市复杂和相似路段问题,提出了基于距离、方向角、连通性、历史数据以及车流方向等约束条件下的地图匹配算法;针对低采样频率浮动车数据,提出了考虑交叉口影响下的行程车速估计算法,通过实测数据评估了算法的性能;最后,针对不同检测技术的特征,提出了数据融合算法.
Intelligent transportation system （ITS） is one effective method of solving traffic problems currently. Furthermore, distribution and prediction of dynamic traffic information are the key problems of ITS. However, because of all kinds of external factors, the traffic condition characterizes in randomicity, which results in great difficulties in distribution and prediction of dynamic traffic information. According to the actuality of a certain city, a kind of brand new dynamic travel speed collection technology based on low frequence FCD is introduced in the paper. First, the empty probe taxi data processing algorithm is brought forward to improve the precision of the techniques. Second, map matching algorithm based on distance, angle, connectivity, as well as historical data is presented to solve the problem of complex and similar sections. Finally, the travel speed estimation and data fusion algorithm is provided, and its precision is verified.
出处
《交通运输系统工程与信息》
EI
CSCD
2008年第4期42-48,共7页
Journal of Transportation Systems Engineering and Information Technology
关键词
浮动车数据
行程车速
聚类分析
数据融合
FCD （Floating Car Data）
travel speed
cluster analysis
data fusion
