数据驱动的疫情应急资源采购-分配鲁棒优化模型  

A Data Driven Robust Optimization Model for Procurement and Allocation of Epidemic Emergency Resource

作　　者：项寅 XIANG Yin(University of Science and Technology of Suzhou,Suzhou,Jiangsu,China)

机构地区：[1]苏州科技大学商学院,江苏省苏州市215000

出　　处：《管理学报》2023年第7期1084-1093,共10页Chinese Journal of Management

基　　金：国家自然科学基金资助项目(72104170);教育部人文社会科学研究规划基金资助项目(21YJC630141)。

摘　　要：为确保应急决策能够根据疫情演化动态调整,建立数据驱动的疫情应急决策框架模型,用滚动决策法将整个疫情期划分为若干离散决策期,根据每个决策期疫情需求的预测结果来优化应急资源采购与分配方案,进一步利用加权最小二乘法修正期末疫情特征信息。针对模型设计求解算法,并结合2020年湖北新冠疫情数据进行算例分析,最终获得关于供应商选择、订单数量决策、应急设施选址和资源分配的集成优化方案,验证了所提出模型与算法的可行性和有效性,得出数据驱动型决策模式相比传统预测-决策型模式更能提高应急效率的结论。To ensure that emergency strategies are dynamically adjusted according to the evolution of epidemics,we propose a data-driven emergency decision model.In this model,the rolling horizon approach is first applied to divide the entire epidemic period into several discrete decision stages,then the optimal procurement and allocation strategies are determined according to the predicted demand.Also at the end of each stage,the least square method is applied to adjust the epidemic information.To solve the model,we developed a novel algorithm.Based on the COVID-19 data of Hubei in 2020,we obtained an integrated optimization plan for supplier selection,order quantity decision,emergency facility location and resource allocation through this model.In addition,the numerical study verifies the feasibility and effectiveness of our model and algorithm,and draws the conclusion that the data-driven model is better than the traditional prediction-decision models.

关 键 词：疫情 应急 采购与分配 数据驱动 鲁棒优化 

分 类 号：C93[经济管理—管理学]