Prediction of urban rail traffic flow based on multiply wavelet-ARIMA model

Abstract

Passenger flow forecast is one of basic reference for the design and operational management of urban rail transit and has become an important procedure in the construction of urban rail transit. In the paper, the ARIMA-Wavelet prediction model was established by analyzing the temporal characteristics of the daily passenger flow change and the principle of the ARIMA model and the Wavelet model, and it was used to forecast the daily traffic flow of Beijing urban rail transit line1. By using the methods of threshold processing, denoising and reconstruction of the original data, it could be better to show the features of general law, then to model and predicate the results by using ARIMA time series model. The test and analysis results of daily prediction of Beijing subway line1 traffic flow, which from November 2 to November 8 in 2015, indicted that Wavelet-ARIMA model was more accurate than ARIMA model. The Wavelet-ARIMA model received better prediction results.