Integrated location-inventory planning of resource warehouses for metro floods: A multi-condition bilevel optimization model

Climate change and urbanization have amplified the threat of storm floods to urban metro systems. During metro flood events, effective and timely transport and deployment of flood control resources is crucial for ensuring the operation of metro stations, which necessitates strategically located and adequately stocked resource warehouses. This study presents a multi-condition bilevel optimization model aimed at integrated planning of the location and inventory for flood control resource warehouses for urban metro systems. In this model, the upper-level decision-making involves developing a location and inventory plan to maximize the weighted average performance of the metro network across various flood conditions. The lower-level decision-making focuses on managing the transport of resources from warehouses to flooded stations based on specific flood conditions and a defined location-inventory plan, with the goal of maximizing passenger flow at protected stations. A case study of the Shanghai metro system validates the model's effectiveness. Results indicate that the model successfully determines optimal warehouse locations, inventory levels, and resource transportation schemes during floods, thereby enhancing metro network performance under various flood conditions. Compared to single-condition optimizations, the multi-condition model yields robust and superior performance of the metro network under floods.