Passenger flow is an important basis for the formulation of transportation organization schemes of a high-speed railway. A reasonable method of passenger flow forecasting can not only accurately predict the passenger flow of railway in the coming years, but also greatly reduce the difficulty and workload of forecasting. Thanks to its adaptive ability and self-learning ability, artificial neural network can be well applied to the passenger flow forecasting of high-speed railways. This study constructed an artificial neural network forecasting model and collected historical data of population, GDP, per capita disposable income, tourism and passenger flow of cities along the section of Shanghai-Kunming High-speed Railway from Nanchang West Station to Changsha South Station for the training of the forecasting model. After comparative analysis with the exponential smoothing method and the time series method, this study used the regression analysis method to forecast the population, GDP, per capita disposable income, tourism and other data of cities along this route, and then forecast the middle and long-term passenger flow between Nanchang West Station and Changsha South Station, hoping to provide reference for the transportation organization of this section. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.
