一种基于修正ARIMA模型的轨道交通站点出站客流预测方法

A Outbound Passenger Flow Forecasting Method Based on Modified ARIMA Model for Rail Transit Station

城市轨道交通客流量预测是其规划、设计、建设和运营各环节的基础和决策依据。该文以具有展会客流特征的地铁站点(以徐泾东站为例)为研究对象,基于交通IC卡等数据,分析展会对相关地铁站点客流量的影响,并建立站点出站客流预测模型。开展了6种常规算法模型的测试和比选,并针对预测结果的"保守"和"滞后",提出修正的ARIMA模型,并给出了预测精度和误差结果,结果显示该模型可有效提高客流预测模型的适用性和精确性。

With the characteristics of passenger flow of the subway stations(in East Xujing station for example) as the research object, based on traffic IC card data, the paper analyzes the influence of exhibition on passenger flow of relevant subway station, and establishes a model for forecasting outbound passenger flow. To carry out the six conventional algorithm models for testing and comparison, according to the prediction results of the 'conservative' and 'lag', the paper puts forward a modified ARIMA model, and gives the prediction accuracy and the error results. The results show that the model can effectively improve the applicability and accuracy of the passenger flow forecasting model.