Optimizing the Distribution of Relief Supplies Considering the Social Vulnerability Risk Index: A Case of Taiwan

Disaster preparedness and the equitable distribution of emergency relief supplies remain pressing priorities for governments worldwide. Situated along the Pacific Ring of Fire, Taiwan is particularly vulnerable to natural disasters that frequently cause severe property damage and endanger public safety, making the efficient allocation of limited relief resources in the immediate aftermath a persistent challenge for emergency management. Although prior research has addressed risk assessment and resource allocation as separate problems, few studies have incorporated region-specific social vulnerability values directly into an optimization framework for relief distribution. This raises the question of how region-specific social vulnerability can be quantitatively embedded into a relief-supply allocation model to achieve a more equitable, risk-sensitive distribution than conventional population-based approaches. To address this gap, the present study develops an integrated decision-making model combining a fuzzy inference system (FIS) with fuzzy multiple objective linear programming (FMOLP). The FIS first derives social risk values from four vulnerability dimensions, namely exposure, disaster mitigation and preparedness, readiness, and recovery, for Taiwan’s 17 counties and municipalities, drawing on multidimensional indicators from the National Science and Technology Center for Disaster Reduction (NCDR) Disaster Mitigation Database. These values are subsequently incorporated as weighting coefficients within the FMOLP model, which jointly maximizes distributional utility and minimizes procurement cost under population-based supply constraints. The results identify Hsinchu, Taichung, Chiayi, Yunlin, and Hualien as the five highest-risk regions. The model further yields a compromise allocation plan spanning all 17 administrative units, with sensitivity analysis confirming its robustness across alternative performance metrics. This study offers a replicable, data-driven decision-support tool to assist disaster preparedness planners and government agencies in relief resource allocation.