Short-term Passenger Flow Prediction for Urban Railway Transit Based on Change-point Model. Issue 1 (October 2020)

• Record Type: Journal Article
• Authors: Yao, Enjian; Chen, Chuanyu; Lu, Muyang; Zhou, Ying
• Is Part Of: IOP conference series. Volume 587:Issue 1(2020)
• Publication Date: 2020-10
• Subjects: Earth sciences -- Periodicals; Environmental sciences -- Congresses; Environmental sciences -- Periodicals; 550.5
• DOI: 10.1088/1755-1315/587/1/012102
• Languages: English
• ISSNs: 1755-1307

Abstract

Abstract: For short-term passenger flow in urban railway transit has the characteristics of nonlinear distribution, a combined forecasting model based on variable-point model and neural networks model is proposed. First, the Pettitt method, the optimal segmentation algorithm, the BG segmentation algorithm(BGSA) and the wavelet analysis method are used to detect the change point of the passenger flow sequence at the memorial hall station, identify the change point of the passenger flow curve and divide the interval, and the passenger flow is predicted and verified by using the Multilayer neural network (MLP) and the radial basis function neural network (RBF). The results show that the wavelet analysis-RBF combined model based on change point detection has the root mean square error of 24.20 and the mean absolute percentage error of 3.52%. Compared with the single neural network model, it improves forecasting accuracy.