灾害应急反应的枢纽集覆盖模型及枢纽最大覆盖模型  

Hub Set-covering Model and Hub Maximal Coverage Model for Disaster Emergency Response

作　　者：陈志宗[1] 关贤军[1] CHEN Zhi-zong GUAN Xian-jun(School of Economics and Management, Tongji University, Shanghai 200092, Chin)

机构地区：[1]同济大学经济与管理学院,上海200092

出　　处：《运筹与管理》2016年第5期15-20,共6页Operations Research and Management Science

基　　金：国家自然科学基金"大规模应急救援资源配送点选址鲁棒优化研究-以地震应急救援为例"(71272047)

摘　　要：重大规模应急管理需要考虑应急反应设施之间的相互连接和支持。基于Campbell对枢纽覆盖的第2类定义,提出了灾害应急反应的枢纽集覆盖模型和枢纽最大覆盖模型。针对枢纽集覆盖模型的过多约束,提出了枢纽集覆盖模型的代理松弛问题,并设计改进的遗传算法对其进行求解。对不同的模型参数与模型规模试验组合的计算结果表明,针对代理松弛问题的改进遗传算法具有计算时间上优势,虽不保证最优解,但多数情况下能得到最优解。Large-scale emerg.ency management needs to consider the interactive links and assistance of emergencyresponse facilities. Based on Campbell＇ s second type of hub coverage definition, this paper proposes the hubset-covering model and hub maximal coverage model for disaster emergency response, which will guarantee boththe travel distance from hubs （emergency response facility）to demand-＇nodes and the discount travel distance be-tween the hubs do not exceed a given coverage radius R. Due to the many constrains of hub set-covering model,this paper also proposes the surrogate relaxation problem of hub set-covering model. Compared to the originalproblem, the surrogate relaxation problem reduces the constraint amounts of（ n ＋ 2） （n - 1 ）/2. We also designthe improved genetic algorithm to solve the surrogate relaxation problem. The computational results in differentexperiment combinations of model parameters and model scale show that the improved genetic algorithm has obvi-ous strengths on the computational time, and the improved genetic algorithm does not guarantee optimal solution,but can get the optimal solution in most cases.

关 键 词：运筹学 枢纽集覆盖 枢纽最大覆盖 灾害应急反应 代理松弛法 遗传算法 

分 类 号：F224.31[经济管理—国民经济]