摘要
为优化应急物资储备库选址,针对传统选址方法忽视功能区划的不足,提出了一种基于两阶段模型的优化选址方法。首先,应用k-means++算法对应急物资需求点聚类,根据聚类结果进行区域划分;然后,基于人口、经济、交通等因素,确定需求点的需求量权重,构建考虑时间成本、覆盖价值和运作成本的选址优化模型。最后,利用非支配排序遗传算法(Non-dominated Sorting Genetic Algorithm Ⅱ,NSGA-)Ⅱ对模型进行求解,得出最优的应急物资储备库选址方案。以H省为例,对其下辖18个市的255个应急物资需求点进行了储备库选址布局优化分析,在保留原有4个应急物资储备库基础上新增4个应急物资储备库。结果表明,采用两阶段优化模型后,储备库选址方案覆盖价值得到了提升,时间成本和运作成本得到了有效降低,为应急物资储备库选址提供了理论参考。
To enhance the site selection process for emergency material warehouses and overcome the shortcomings of traditional methods that fail to consider functional zoning,this paper introduces a two-stage optimization approach.Initially,we surveyed emergency material demand points to establish their coordinates.After evaluating various clustering algorithms,we selected the k-means++algorithm due to its exceptional clustering performance,as indicated by the silhouette coefficient and elbow method analyses,to cluster the demand point coordinates effectively.The clustering results were utilized for regional division,providing a foundation for functional zoning.Subsequently,we assessed factors including population,economic conditions,and transportation to determine the weighted demand for emergency materials at each demand point.Taking multiple considerations into account,we identified three key objectives:time cost,coverage value,and operational cost.A multi-objective site selection optimization model was then developed,integrating these three dimensions.The Non-dominated Sorting Genetic Algorithm Ⅱ(NSGA-Ⅱ)multi-objective optimization algorithm was utilized to solve the model and identify optimal site selection solutions.Using Province H as a case study,the k-means++algorithm was applied to 255 emergency material demand points across 18 cities.The results indicated that the province was effectively divided into five regions,closely aligning with its functional zoning plan.This outcome demonstrates the efficacy of the clustering algorithm in facilitating functional zoning.Considering the weighted demand at each point and the three objectives—time cost,coverage value,and operational cost—NSGA-Ⅱ was employed to generate seven optimized site selection schemes for ten candidate warehouse locations.Among these,the optimal scheme retained eight warehouses,comprising four existing ones and four newly established emergency material warehouses.The results demonstrate that the two-stage optimization model successfully achieves an effective trade-off among multiple objectives,enhancing coverage value while significantly reducing both time and operational costs.This provides valuable theoretical guidance for the site selection of emergency material warehouses within the framework of public safety.
作者
张召冉
张楠
杨明
ZHANG Zhaoran;ZHANG Nan;YANG Ming(College of Civil Engineering,North China University of Technology,Beijing 100144,China;Beijing Beikai Electric Co.Ltd.,Beijing 100176,China)
出处
《安全与环境学报》
北大核心
2025年第8期3189-3197,共9页
Journal of Safety and Environment
关键词
公共安全
应急物资库
选址优化
功能区划
多目标规划
public safety
emergency material warehouse
site selection optimization
functional zoning
multi-objective planning
作者简介
张召冉,副教授,从事工程技术施工与安全管理研究,2415661256@qq.com。
