A Clustering-based Location Allocation Method for Delivery Sites under Epidemic Situations

基于聚类的疫情物资投放点选址方法研究

To address the poor performance of commonly used intelligent optimization algorithms in solving location problems—specifically regarding effectiveness,efficiency,and stability—this study proposes a novel location allocation method for the delivery sites to deliver daily necessities during epidemic quarantines.After establishing the optimization objectives and constraints,we developed a relevant mathematical model based on the collected data and utilized traditional intelligent optimization algorithms to obtain Pareto optimal solutions.Building on the characteristics of these Pareto front solutions,we introduced an improved clustering algorithm and conducted simulation experiments using data from Changchun City.The results demonstrate that the proposed algorithm outperforms traditional intelligent optimization algorithms in terms of effectiveness,efficiency,and stability,achieving reductions of approximately 12%and 8%in time and labor costs,respectively,compared to the baseline algorithm.

为解决常用的智能优化算法在求解选址问题时,在有效性、效率和稳定性方面表现不佳的问题,提出了一种新的在隔离期间物资投放点的选择方法。在确定优化目标和约束条件后,根据所收集的数据建立了相关数学模型,并采用传统的智能优化算法得到Pareto前沿解;基于这些Pareto前沿解的特点,提出了一种改进版聚类算法,并利用长春市的相关数据进行了仿真实验。结果表明:所提算法在有效性、效率和稳定性方面均优于传统的智能优化算法,与基准算法相比在时间成本和人力成本上分别降低了约12%和8%。