The application of ARIMA-RBF model in urban rail traffic volume forecast

Due to various factors, passenger flow has nonlinear characteristics. Autoregressive Integrated Moving Average Model (ARIMA model) is suitable for non-stationary time series forecasting while RBF neural network is a kind of forward neural network which has good approximation performance and is suitable for processing nonlinear problem. In this paper, we combine the ARIMA model and RBF neural network model to formulate the ARIMA RBF model by analyzing passenger flow’ s temporal characteristics, the mechanism of ARIMA model with RBF model. We use the proposed model which used to forecast Beijing urban rail transit passenger flow and obtain a good prediction result. KeywordsRailway traffic; Passenger flow forecast; Combination Forecasting; RBF neural network; ARIMA model