基于历史数据和时间系数的城市轨道站点客流预测

Prediction of Urban Rail Station Passenger Flow Volume Based on Historical Data and Time Coefficient

在对城市轨道交通站点客流规律、特点进行分析的基础上,通过对上海市轨道交通系统客流历史数据的深入分析,在引入平假日系数的同时,提出了小时系数的概念,并建立了基于平假日系数和高峰小时系数的城市轨道交通站点客流神经网络预测模型,提高了站点进站客流预测的精度,为城市轨道交通系统客流管理提供了理论基础.

The characteristics of urban rail transit passenger flow volume was analyzed in-depth through the historical data.While the flat holiday coefficient was introduced,a concept of hour coefficient was proposed.Based on historical data,the flat holiday and hour coefficient,this paper established a neural network prediction model to improve the accuracy of urban rail transit site passenger flow forecast,and provided the theoretical basis for passenger flow management of urban rail transit system.