基于云模型的城市轨道交通短时客流预测

Forecast of Short-term Passenger Flow in Urban Rail Transit Based on Cloud Model

城市轨道交通线路短时客流具有不确定性特征。分析了短时客流的准周期性,用云概念表示短时客流的特征,构建历史时间云、历史客流云、当前客流趋势云以及客流预测云,并建立时间云与客流云的关联规则,将时间云作为规则前件,客流预测云作为规则后件构建单条件多规则不确定性预测云模型。以南京地铁2号线15 min间隔的进站客流预测为例,将云模型与ARIMA模型的预测结果进行对比分析,证明云模型应用于短时客流预测的有效性,从而为城市轨道交通线路短时客流预测提供了一种新途径。

The short-term passenger flow of urban rail transit has uncertain characteristics.Based on an analysis of quasi-periodical of short-term passenger flow,the basic theory and methods of cloud model are used to represent the short-term passenger flow characters,and build the passenger flow cloud,time cloud,historical time cloud,historical passenger flow cloud,current passenger flow trend cloud and the passenger flow forecasting model.An uncertainty forecast model with single conditional and multiple rules is formed by setting the time cloud as the former and the passenger flow forecast cloud as the latter in forecast rules.At the same time,the model is used to forecast the entrance passenger flow of 15 min time interval in Nanjing Metro Line 2,and compared with ARIMA model,to prove that cloud model is effective and can provide a new way for the short-term passenger flow forecast in urban rail transit.