Short-term inbound passenger flow prediction for urban rail transit by integrating spatiotemporal features

Accurate subway passenger flow prediction is crucial for intelligent transportation systems to manage traffic, optimize operations, and plan infrastructure. Since passenger flow is affected by factors like station location and date type, further research on short-term high-frequency prediction is needed. This paper proposes a method combining Graph Convolutional Networks (GCN) for spatial dependencies and Informer models for long-term temporal dependencies to capture spatiotemporal relationships. By fusing these features in a fully connected layer, the approach achieves precise passenger flow forecasts. Experiments on Hangzhou Metro Line 1 show that this method is robust and accurate.