基于浮动车数据的城市道路行程时间估计

Travel Time Estimation for Urban Roads Based on Floating Car Data

为提高城市道路行程时间估计模型的准确度和有效性,本文利用浮动车数据,依据对传统模型思路的总结分别建立了基于路段长度比例和点速度调和平均值的两种行程时间初阶估计模型,并利用统计学中的同分布融合思想建立了行程时间融合模型,以修正初阶模型结果的精度,弥补传统估计模型中准确度低、效率不高的缺陷。最后选取成都市具有代表性的路网区域为算例,验证了初阶模型假设分布的正确性,同时计算出融合模型路径总时间的平均偏差仅为12%,说明了融合模型的准确度和有效性。

To improve the accuracy of the model for estimating travel time of urban roads, based on we develop two kinds of initial estimation models of travel time based on the length ratio of the road and the mean value of the point velocity respectively. The two models employ the floating car data and the traditional model idea. Then a travel time fusion model is developed based on the same-distribution fusion ideas in statistics, which corrects the accuracy of the initial estimation models. This also improves the accuracy and efficiency of the tradition models. Finally, a representative road network area in Chengdu is selected as an example. The correctness of the hypothetical distribution of the initial estimation models are verified and the average deviation of the total path time of the fusion model is calculated to be only 12%, demonstrating the accuracy and validity of the fusion model.