基于贝叶斯网络的城市应急物资储备库选址多目标优化  

Multi-objective Optimization of the Reserve Site Selection for Urban Emergency Supplies Based on the Bayesian Network

作　　者：王雷 WANG Lei(Henan Technical Institute,Zhengzhou,Henan 450042,China)

机构地区：[1]河南应用技术职业学院,河南郑州450042

出　　处：《物流技术》2025年第3期61-72,共12页Logistics Technology

基　　金：2025年度河南省高等学校重点科研项目“移动边缘计算赋能智能网联车的多维资源融合协作决策与优化调度研究”(25A413012)。

摘　　要：为了提升城市应急物资储备库选址方案的资源分配均衡度,提出了一种基于贝叶斯网络的城市应急物资储备库选址多目标优化方法。基于城市应急物资储备库的分布情况,以最小化响应时间、最大化物资覆盖区域、最小化选址成本为优化目标,建立储备库选址的多目标函数,平衡各个优化目标在选址过程中的重要度。在考虑储备库及应急物资总量的前提下,对多目标函数设定限制条件,确保应急物资的储备量和分布能够满足快速响应的需求。通过引入贝叶斯网络,将储备库选址过程中的所有不确定因素作为贝叶斯网络的节点,确定节点的先验概率和条件概率,动态更新多目标函数,生成最优的储备库选址方案。并以河南省郑州市某区域为例进行实验分析,结果表明,优化后的选址方案资源分配均衡度均值提升至0.94,显著优于其他对比算法,且在多个突发事件场景下表现出较高的稳定性和适应性。This paper proposes a multi-objective optimization method based on the Bayesian network for enhancement of the resource allocation balance in the reserve site selection for urban emergency supplies.Given the distribution of the reserves,a multi-objective function is constructed to balance the importance of each optimization target in the site selection process,covering minimization of the response time,maximization of the material coverage area,and minimization of the site selection costs.Under the premise of the total number of reserve sites and the quantity of emergency supplies,the multi-objective function is subject to additional constraints to ensure that the reserves and their distribution of emergency supplies can meet the demands for rapid emergency response.By introducing the Bayesian network,the multi-objective function is dynamically updated to generate the optimal site selection scheme,as all uncertainties in the site selection process are modeled as nodes with prior and conditional probabilities defined for each.A case study,conducted in a specific area of Zhengzhou,Henan Province,demonstrates that the optimized scheme improves the average resource allocation balance to 0.94,significantly outperforming other comparative algorithms,and shows strong stability and adaptability under various emergency scenarios.

关 键 词：贝叶斯网络 城市应急物资 储备库选址 多目标优化 资源分配均衡度 

分 类 号：F251[经济管理—国民经济] O221[理学—运筹学与控制论]