低频GPS数据和交叉口延误下的行程时间估计

为解决低频GPS数据条件下路段行程时间估计误差较大的问题,分析了车辆在道路交叉口影响区域的延误特征,根据两个相邻GPS点跨越一个或多个交叉口,以及跨越的交叉口影响区域內有无GPS点,划分了4种GPS分布类型,并设计了相应类型的基于交叉口延误计算的路段行程时间插值算法.以北京市八角地区实际GPS数据为例验证了本文算法,结果表明:在平峰和高峰时段,用本文算法计算的路段行程时间的平均绝对相对误差不超过14%;与改进插值法相比,平均估计精度提高了7.9%,在交叉口延误时间较大的路段效果更显著. 

Link Travel Time Estimation for Urban Roads Using Low Frequency GPS Data and Intersection Delay

In order to solve the problem of large error in link travel time estimation with low-frequency GPS data, vehicle delay features in road intersection-affected areas were analyzed. According to whether two adjacent GPS points stride over multiple intersection-affected areas and whether there are GPS points in the intersection-affected areas, GPS distribution types were divided into four types and a difference algorithm for travel time estimation was designed based on calculation of intersection delay time. This travel time estimation algorithm was then verified using the vehicle GPS data in the Bajiao area of Beijing in a case study. The results show that the average absolute errors of link travel time estimates by the algorithm are within 14% in non-peak and rush hours. Compared with an improved interpolation method, the proposed algorithm improves the average estimate accuracy by 7.9%, and the effect is more obvious in the link with a larger intersection delay time.