Research on Train Delay Forecast of High-Speed Railway Based on Data Driving

China’s high-speed railway exhibits high train density and small tracking intervals. Once a train is delayed, the delay easily propagates, leading to further delays. Therefore, accurate prediction of train delays can assist dispatchers in formulating timely optimization and adjustment plans, minimizing the negative impact of delay propagation, and enhancing the reliability and punctuality of train operations. This paper constructs a train delay prediction model based on the cyclic neural network (RNN) algorithm, enabling accurate prediction of train arrival delay times. The model is then validated using actual operational data from the Wuhan-Guangzhou high-speed railway, with the experimental results demonstrating high prediction accuracy.