Passenger flow prediction of subway stations based on ARIMA-RBF
By analyzing the fundamental principles of the ARIMA model and the RBF model,and combining the time series forecasting characteristics of ARIMA with the nonlinear processing capabilities of the RBF neural network,an ARIMA-RBF hybrid model is developed to predict passenger flow at subway stations.Finally,using the Shiergong Station on Guangzhou Metro Line 2 as a case study,the ARIMA-RBF hybrid model is applied to real-world passenger flow prediction scenarios.The results show that the ARIMA-RBF hybrid model achieves an error rate of just 2.23％,which is superior to that of the traditional ARIMA model.
station passenger flowARIMA modelRBF modelARIMA-RBF combined prediction
万方数据
维普
