Comparison of Several Traffic Forecasting Methods Based on Travel Time Index Data on Weekends

Traffic forecasting provides the estimation of future traffic state to help traffic control,travel guide,etc. This paper compared several widely used traffic forecasting methods,and analyzed each one's performance in detail to make conclusions,which could redound to researchers choosing an appropriate traffic forecasting method in their own works. Compared with conventional works,this paper creatively assessed the performance of traffic forecasting methods based on travel time index (TTI) data prediction,which made the accuracy of our comparison better.

基于高速公路收费数据的路网运行状态评价

探讨利用高速公路收费数据作为数据资源进行路网运行状态评价的可行性,并完整分析了相关算法。选取某实际运营的高速公路网2008—2010年间7～10月的收费数据作为基础,重点介绍数据的分析处理过程,包括进行数据准备,执行各类数据质量控制规则、资料库配置及构建样本输出模型等,深入挖掘高速公路网运行存在的出行时间分布及周期分布等潜在规律,计算和提取了平均速度、出行延误、行驶时间指数、计划行驶时间指数、拥堵持续时间等几类路网运行状态评价的核心指标,并指出了采用高速公路收费数据作为数据资源开展此类研究的优缺点。分析表明,收费数据可应用于高速公路网运营绩效的评估,并可为交通出行提供参考信息。