在突发公共卫生事件时，最优的生活物资发放位置的选择极为关键，为此，以长春市的应急生活物资发放为例进行实证研究，提出应急物资发放选址优化算法。首先，基于K-means聚类和泰森多边形的方法确定虚拟物资仓库、发放备选点以及服务覆盖区域。然后，以最大覆盖率、最小化成本为目标建立优化模型，并提出基于自适应精英保留策略的改进遗传算法进行求解。实证研究表明聚类分析算法能够初步得到873个虚拟物资仓库和发放备选点，而泰森多边形能够有效确定备选点的合理覆盖范围，改进的遗传算法能够得到更优的30个物资仓库和发放位置。数据分析同时发现从物资发放到有效控制传染存在时间滞后性，因此一旦新增速率超过正常值，应该提前加大应急物资储备和发放以控制感染的进一步爆发。研究成果有利于增强城市危机管理和应急响应的能力，确保生活物资供应的及时性和高效率。

In the event of a public health emergency, the selection of the optimal distribution locations of daily necessities is crucial. Accordingly, the emergency distribution of daily necessities in Changchun City was taken as an example for empirical research and an optimization algorithm was proposed for the location of emergency supplies distribution. Initially, the virtual material warehouses, distribution candidate locations, and service coverage areas were determined by using K-means clustering and the Thiessen polygon method. Then, an optimization model was established with the goals of maximum coverage and minimum cost, and an improved genetic algorithm based on adaptive elite retention strategy was proposed to solve it. Empirical research had shown that the clustering algorithm could obtain 873 virtual material warehouses and distribution candidates, while Thiessen polygons could determine the reasonable coverage range of candidates, and improved genetic algorithms could obtain better 30 material warehouses and distribution locations. Data analysis had also found that there was a time lag between the distribution of supplies and effective control of infection. Therefore, once the rate of new additions exceeds the normal value, emergency supplies should be increased in advance to control further outbreaks of possible infections. The research results are beneficial for enhancing the ability of urban crisis management and emergency response, ensuring timely and efficient supply of daily necessities.