Passenger Redistribution Based on Rolling Horizon Optimization in Multimodal Emergency Evacuation of Transportation Hubs

Passenger emergency evacuation is a critical aspect of operations at multimodal transportation hubs. When a large number of passengers gather within the hub, it is necessary to evacuate them efficiently from multimodal transfer points to ensure their safe departure and to prevent unforeseen incidents such as stampedes. To improve the efficiency of emergency evacuation under limited evacuation capacity conditions, passenger redistribution can be adopted as a strategy to optimize the evacuation process. This study proposes a passenger redistribution model based on a rolling horizon optimization framework for the emergency evacuation of multimodal transportation hubs. The model is formulated as a multiobjective optimization problem with two objectives: (1) minimizing the total evacuation time to maximize the overall evacuation efficiency and (2) minimizing differences in waiting times among different travel modes to reduce imbalance across them. The problem is solved using the NSGA‐II algorithm. A case study based on Terminal 3 of Beijing Capital International Airport is conducted to demonstrate the effectiveness of the proposed passenger redistribution model. The results show that the proposed model can generate effective passenger redistribution plans for emergency evacuation of the transportation hubs. Furthermore, the findings indicate that passenger redistribution is particularly suitable when there is a significant mismatch between evacuation capacity and evacuation demand. However, its effectiveness is limited when a large number of passengers are initially stranded, while it proves to be more effective when the initial number of stranded passengers is relatively small.