Forecasting Freight Flows and Mitigating Disruption Risks in the China–Europe Multimodal Transport Corridor: A Physics/Economics-Informed Spatiotemporal Diffusion Bi-LSTM Approach

We develop ST-Diff-BiLSTM to forecast node-level freight flows and classify disruption risk in the China–Europe multimodal corridor. The model combines geographic, trade, and flow-based graph links with feature and temporal attention. Its loss includes capacity and flow-balance penalties. In the reported test split, MAE was 8.4 TEU, compared with 17.8 TEU for baseline LSTM; the capacity-overrun rate was 0.1%, compared with 4.8%. The disruption classifier reached an AUC of 0.941. These results describe the reported dataset and simulation settings; broader operational performance requires external validation.