计划性大型活动散场期间地铁OD客流量预测方法

Prediction of Subway OD Passenger Flow during the end of Planned Large-scale Events

在计划性大型活动举办期间地铁客流量会超过平常高峰期的客流量.客流的激增不仅对城市轨道交通正常运营造成巨大的压力,甚至会引起严重的安全事故.本文在对计划性大型活动散场期间地铁客流在时空范围的规律研究基础上,根据大型活动散场时段的OD客流量基础数据,结合影响因素特征数据,构建基于随机森林算法的计划性大客流预测模型,实现大型活动结束后在5 min粒度下的OD客流量预测,并以北京凯迪拉克中心五颗松地铁站为例进行实例研究.选取演唱会和体育赛事的AFC数据,对预测结果进行验证,并与SVM、XGBoost算法对比,证明本文所提出的基于随机森林算法的客流预测模型方法具有更好的预测效果.

In the daily operation of urban rail transit,the passenger flow during the end of planned large-scale events will greatly exceed normal passenger flow,which not only causes great pressure on the normal operation of urban rail transit,but even leads to serious safety issues.This paper first analyzes the characteristics of subway passenger flow in the space and time range during the closing of planned large-scale events.Then,on the basis of massive origin_destination traffic flow data and the characteristic data of influencing factors,a planned large-scale passenger flow prediction model based on random forest algorithm is estimated.This model can predict the origin_destination traffic flow at a granularity level of 5 minutes after the large-scale event.This paper takes the Wukesong Station of the Beijing Cadillac Center as an example and collects the AFC data of concerts and sports events to verify the prediction results.The results show that the passenger flow prediction model method based on the random forest algorithm proposed in this paper has a better prediction effect after being compared with the SVM and XGBoost algorithms.