Abstract:

Sudden natural disasters present substantial challenges that require swift and efficient response strategies. To adequately prepare for these events， it's essential to build emergency facilities and stockpile relief supplies in these facilities in advance. However， despite the significant role household reserves play in mitigating the impacts of unforeseen natural disasters， existing research on emergency facility location and relief supply distribution has not examined the impact of household reserves on decision-making. To address this research gap， an emergency facility location and relief supply distribution problem considering household reserves in the context of earthquake disasters is investigated. Before a disaster occurs， strategic decisions are made regarding the location and capacity of emergency shelters and relief warehouses， the amount of relief supplies pre-positioned， and the selection of areas for household reserves. Subsequently， tactical decisions for evacuating residents and distributing relief supplies are determined once a disaster occurs. The uncertainties encompassing disaster characteristics， relief supply， transportation and demand are considered. To make reliable decisions amid these uncertainties， a novel two-stage distributionally robust optimization model is introduced. This model aims to minimize the total expected costs of rescue in the worst case， including direct costs from pre-disaster preparation and post-disaster response， as well as penalty costs for unmet demands.To solve the proposed model effectively， a two-phase column-and-constraint generation （C&CG） algorithm is developed. The effectiveness of the proposed model and algorithm is verified through a case study based on historical earthquakes in a region of Yunnan Province， China. By highlighting the crucial role of household reserves in the emergency facility location and relief supply distribution problem， a novel perspective on emergency management is offered. In addition， the introduction of a two-stage distributionally robust optimization model offers a comprehensive framework for improving disaster preparedness and response strategies， providing a valuable tool for policymakers and emergency management practitioners. Based on the results of the case study， some important managerial insights are drawn as follows： Firstly， emergency management practitioners should promote the notion of self-responsibility among residents for their safety while ensuring their access to relief supplies after a disaster. Secondly， guiding and training residents to properly use household reserves to increase the post-disaster availability of supplies plays a crucial role in reducing disaster impact on people and lowering government expenses. Lastly， for the problem of emergency facility location and relief supply distribution， decision-makers need to balance the use of available data with the inherent uncertainties of such unpredictable disasters.

Key words: emergency facility location, household reserves, robust optimization, C&CG algorithm