基于分布式鲁棒的城市群救灾网络与灾民疏散联合规划研究  

A distributionally robust approach for joint planning of relief network and victim distribution based on regional coordination of urban agglomeration

作　　者：段倩倩 张梦玲 白鹏飞 黄钧 DUAN Qian-qian;ZHANG Meng-ling;BAI Peng-fei;HUANG Jun(School of Logistics,Beijing Wuzi University,Beijing 101149,China;School of Management and Economics,Beijing Institute of Technology,Beijing 100081,China;School of Management and Engineering,Capital University of Economics and Business,Beijing 100070,China;School of Engineering and Science,University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]北京物资学院物流学院,北京101149 [2]北京理工大学管理与经济学院,北京100081 [3]首都经济贸易大学管理工程学院,北京100070 [4]中国科学院大学工程科学学院,北京100049

出　　处：《安全与环境学报》2022年第3期1450-1458,共9页Journal of Safety and Environment

基　　金：国家社会科学基金项目(17BGL176)。

摘　　要：城市群的出现极大地提高了城市间的分工协作功能,大规模灾害发生时,其对应急资源的高效配置和应急救援活动的快速展开至关重要。为了提高城市群的应急救援效率,基于城市群区域协同机制,进行了应急设施的选址、灾民疏散和物资配送联合规划研究。大规模灾害发生之前,通过构造一个包含多个指标因素的综合评价体系,对各应急设施面对灾害的应急响应时间进行分区处理,确定设施选址及其差异化的应急响应能力;灾害发生之后,依托避难所和配送中心完成各灾区灾民的安置和物资的需求满足。由于地震灾难发生具有不确定性的特点,引入联合机会约束的方法来确保各灾区灾民疏散的可靠性,以此建立了总成本最小化的分布式鲁棒优化模型,并将其转化为易于求解的二阶锥规划模型。结果表明,联合机会约束的分布式鲁棒模型在选址数量上少于随机模型,并以较少的成本疏散了更多的灾民,相较于传统鲁棒模型,降低了其保守性。同时分析了城市群的区域异质性,表明考虑区域异质性使得整个救援过程更具针对性、有效性和经济性。The emergence of urban agglomerations has greatly improved the functions of division of labor and cooperation between cities.Previous studies focus on emergency relief activities without considering the regional coordination mechanism of urban agglomerations.Furthermore,most studies discuss the location of emergency facilities,evacuation of victims,and emergency material distribution separately.To improve the emergency rescue efficiency of urban agglomerations,we study the location of emergency facilities,evacuation of victims,and emergency material distribution based on the regional coordination mechanism of urban agglomerations.Before large-scale disasters occur,we need to construct a system that contains multiple indicator factors.The comprehensive evaluation system for emergency facilities is used to deal with the speed of emergency response time in the face of disasters and determine their location and their differentiated emergency response capabilities;after the disaster occurs,the refuge and distribution center is used to complete the resettlement of the victims in each disaster area and thus meet the demand for materials.Due to the uncertainty of the occurrence of earthquake disasters,this paper introduces a joint chance constraint method to ensure the reliability of the evacuation of victims in affected areas.We develop a distributionally robust optimization model with the minimum total cost and transform it into a computationally tractable second-order cone programming model.Finally,the real-world case study illustrates the efficiency of the model and provides managerial insights.The results show that the distributionally robust model with joint chance constraints is less than the stochastic model in terms of the number of locations,and evacuates more victims at a lower cost.Compared with the traditional robust model,it reduces its conservativeness.Moreover,considering the regional heterogeneity makes the entire rescue process more targeted,effective,and economical.

关 键 词：公共安全 城市群 疏散配送 联合机会约束 分布式鲁棒优化 

分 类 号：X921[环境科学与工程—安全科学]