一种基于历史数据与实时信息的公交路段行程时间预测

Prediction of Travel Time Based on Historical Data and Real Time Information

在分析公交车的运行过程及影响因素基础上,根据公交车行驶的周期性特点,提出了根据行驶中公交车过站时刻的历史数据分类方式,为了反映当前的行驶状况,引入了最近邻车次在目标区段即时行驶信息对历史数据预测结果进行调节。实验结果表明,此种预测方法与典型的回归分析方法相比减少了需要的数据量和复杂度,与历史数据统计法相比对偶发性情况有更好的适应性。

On the basis of analyzing the running process and the influencing factors of the bus,the cyclical characteristics of the bus running are pointed out.A data classification method based on the arrival time is proposed in order to reflect the current travel situation.The same serial number buses driving informations on the trips recently between the current site and the target site are used to adjust forecast data.Experimental results show that this method reduces data amount and data complexity compared with the typical regression analysis method,and it has better adaptability compared with the method of historical data statistics.