Study on Spring Festival Passenger Flow Forecasting of Rail Transit in South China Based on Optimized SARIMA Model--Taking Guilin City as an Example

In this study, the optimized SARIMA model is used to forecast the spring rail transit passenger flow in Guilin City in 2019-2025, and external factors such as GDP and mobile population are introduced to improve the forecasting accuracy. First, the data range is extended to 11 years to enhance the model learning ability and reduce the impact of epidemics. Subsequently, the SARIMA parameters are optimized using grid search and AIC criterion to ensure the optimal fitting effect, and the prediction stability is improved by error control strategy. The experimental results show that the optimized SARIMA model performs well in short-term prediction, and the prediction error decreases year by year. Based on the prediction results, strategies such as mobility management, ticket optimization and capacity allocation are proposed to alleviate the pressure of passenger flow during the Spring Festival. This study provides a scientific basis for rail transportation capacity planning, which is of great significance to improve the transportation efficiency during the Spring Festival.