摘要
大型突发事件发生后需要快速启动应急救灾网络,合理配置应急医疗服务站。本文考虑各应急医疗服务站选址节点需求的不确定性,引入三个不确定水平参数,构建四类不确定需求集合(box,ellipsoid,polyhedron和interval-polyhedron)对应的应急医疗服务站鲁棒配置模型,运用分支-切割算法求解,最后,进行需求扰动比例的灵敏度分析。算例结果表明,四类不确定需求集下的鲁棒配置模型中,ellipsoid不确定需求集合配置模型开放设施较少,总成本最小,鲁棒性较好。决策者还可以根据风险偏好选择不确定水平和需求扰动比例的组合,以使得总成本最小。
After large-scale emergent incidents, it is particularly urgent to start emergency relief network quickly and locate the emergency medical service stations. We incorporate demand uncertainty, introduce three uncertain level parameters, construct four types of demand uncertainty sets, and propose robust EMSS location models respectively. Then, we derive their tractable robust counterparts, and employ branch-cut algorithm to solve the problems. Finally, we present computational results, and perform sensitivity analysis. Numerical results show that, among four robust location models, robust EMSS location model with ellipsoid uncertainty set opens fewer EMSSs, and the total cost is smaller. According to the uncertainty of risk preference, decision-makers usually choose optimal budget uncertainty and demand disturbance proportion, so as to minimize the total cost.
出处
《运筹与管理》
CSSCI
CSCD
北大核心
2017年第9期21-28,共8页
Operations Research and Management Science
基金
国家自然科学基金资助项目(71432002
71172172
71272058)
关键词
应急医疗服务
设施选址
鲁棒优化
不确定需求集合
鲁棒等价
emergency medical service
facility location
robust optimization
demand uncertainty sets
robust counterpart
