流域性洪灾动态演进下的应急物资调度模型  

Model of emergency material scheduling under dynamic evolution of river basin flood

作　　者：杨泽鸿 白雨凡 闫兵艳 段在鹏 YANG Zehong;BAI Yufan;YAN Bingyan;DUAN Zaipeng(College of Environment and Safety Engineering,Fuzhou University,Fuzhou Fujian 350108,China;School of Economics and Management,Fuzhou University,Fuzhou Fujian 350108,China;Fujian Emergency Management Research Center,Fuzhou Fujian 350108,China)

机构地区：[1]福州大学环境与安全工程学院,福建福州350108 [2]福州大学经济与管理学院,福建福州350108 [3]福建省应急管理研究中心,福建福州350108

出　　处：《中国安全生产科学技术》2025年第4期47-56,共10页Journal of Safety Science and Technology

基　　金：国家社会科学基金项目(23BGL290)。

摘　　要：为解决流域性洪灾多灾点动态调度问题,首先通过案例推理确定物资出救量以保证灾后快速出救,其次通过贝叶斯网络进行洪水演进发展过程中灾情风险动态分析,之后根据救援阶段分别以物资满足率、时效性和以物资满足率、调度成本、灾后风险损失为优化目标构建模型,最后通过改进的多目标遗传算法对多物资、多灾点、多救援点的多目标模型进行求解,获得各阶段最优调度分配方案。研究结果表明:模型以流域性洪灾为研究对象,在分析多个灾点的动态风险变化后,对救援各阶段采取不同的目标,最终得到的分配方案能够在灾害初始阶段考虑时效性的同时达到87.13%的综合满足率,在后续演进发展阶段中考虑风险因素,能够在控制成本的同时提高各灾点满足率,可有效解决流域性洪灾的物资供需问题。In order to solve the problem of dynamic scheduling for multiple disaster points in river basin flood,firstly,the amount of material rescue was determined by the case-based reasoning to ensure the rapid rescue after the disaster.Secondly,the dynamic analysis of disaster risk in the process of flood evolution and development was carried out by Bayesian network.Then,according to the rescue stage,the models were constructed with the material satisfaction rate,timeliness and the material satisfaction rate,scheduling cost and post-disaster risk loss as the optimization objectives,respectively.Finally,the improved multi-objective genetic algorithm was used to solve the multi-objective model of multiple materials,multiple disaster points and multiple rescue points,and the optimal scheduling allocation scheme of each stage was obtained.The results show that the model takes the basin flood disaster as the analysis object,analyzes the dynamic risk change of multiple disaster points,and adopts different objectives in each stage of rescue.The final allocation scheme can achieve a comprehensive satisfaction rate of 87.13%while considering the timeliness in the initial stage of disaster.Considering the risk factors in the subsequent evolution and development stage,it can improve the satisfaction rate of each disaster point while controlling the cost,and effectively solve the problem of material supply and demand in river basin floods.

关 键 词：流域性洪灾 多受灾区域 风险演变 物资调度 多目标遗传算法 

分 类 号：X913[环境科学与工程—安全科学]