基于时间序列的趋势外推模型预测城市轨道交通车站客流的应用  

Application of trend extrapolation model based on time series in passenger flow prediction of Urban Transit station

作　　者：王超[1] 钱进[1] 李军[2] 赵静[1] 

机构地区：[1]中国铁道科学研究院电子计算技术研究所,北京100081 [2]广州市地下铁道总公司,广州510030

出　　处：《铁路计算机应用》2012年第5期50-51,55,共3页Railway Computer Application

摘　　要：本文提出了一种基于时间序列的趋势外推模型,对实际应用中的客流数据进行预测,通过预测与节假日客流数据变化规律的分析,对模型进行完善与修正。在地铁日常运营操作过程中,起着辅助判断的作用,并能及时预判问题采取措施进行预防。该算法在城市轨道交通应急平台项目中应用,解决了线路车站日常客流运营维护中出现的大客流事件发生的问题。This paper presented a method based on trend extrapolation model of time series to predict the data of passenger flow in practical application. The model was perfected and corrected through predicting the holiday passenger flow data and analyzing the changing rule of the data. In the daily operation of the metro, the model played an auxiliary role of judgment to take timely pre-judgment measures for prevention. The algorithm mentioned in this paper firstly used in emergency platform project of Guangzhou Metro, solved the problems of large passenger flow data occurred in the daily operation and maintenance of passenger station.

关 键 词：客流预测 趋势外推 客运站 

分 类 号：U231[交通运输工程—道路与铁道工程] TP39[自动化与计算机技术—计算机应用技术]