城市轨道交通客流预测算法研究

Research on Passenger Flow Prediction Algorithm of Urban Rail Transit

城市轨道交通客流预测是客运组织的基础,预测结果可为运营管理提供决策依据.针对城市轨道交通客流量预测问题,提出了一种ARMA-RBF组合客流预测算法:首先根据变点算法,通过对客流数据构成的时间序列处理得到变点集;然后基于小波变化对变点集进行去噪处理;最后利用ARMA-RBF算法对城市轨道客流进行预测.以北京地铁4号线新街口、平安里、西四地铁站进出客流数据进行方法验证,结果表明,相较于单独的ARMA算法或RBF神经网络算法,ARMA-RBF组合客流预测算法可提高城市轨道交通进站客流预测的精度.

The passenger flow prediction of urban rail transit is the basis of passenger transport organization,and the prediction results can provide a decision-making basis for operation management.Aiming at improving the urban rail transit passenger flow prediction,this paper proposes an ARMA-RBF combined passenger flow prediction algorithm.First,according to the change point algorithm,the change point set is obtained by processing the time series of passenger flow data;second,the change point set is denoised based on the wavelet change;finally,the urban rail passenger flow is predicted by using the ARMA-RBF algorithm.The method is verified by the passenger flow data of Xinjiekou,Pinganli and Xisi subway stations of Beijing line 4.The results show that,compared with the single ARMA algorithm or the RBF neural network algorithm,ARMA-RBF combined with the passenger flow prediction algorithm can improve the accuracy of urban rail transit inbound passenger flow prediction.