数据与SEIR模型双驱动下COVID-19疫苗的空间分配策略

Data-and-SEIR-model dual-driven spatial allocation strategies of COVID-19 vaccines

为提出一个适合疫情中后期新冠疫苗空间分配的新框架,通过收集2022年11月13日至28日国内31个省会城市的新冠疫情数据,求解传染病模型的逆问题得到初始潜伏人数,借助K-means算法对城市风险等级进行分类,进而根据风险分级后的城市建立疫苗分配模型,并通过遗传算法求解得到最优分配策略。数据结果表明,31个省会城市可分为5个疫情风险等级区域,风险等级越高的地区,疫苗在待接种人群的覆盖比例越高;在降低累计感染人数方面,所提出的分配策略优于地区人数比例分配策略和风险等级优先全覆盖分配策略。

In order to present a new framework for spatial allocation of COVID-19 vaccines in the middle and late stages of the epidemic,we collected the epidemic data of 31 provincial capitals in China from November 13 to 28,2022,solved the inverse problem of the epidemic model,and classified cities into different risk levels by the K-means algorithm.The vaccine allocation model is established based on risk classification,and the optimal allocation strategy is obtained by genetic algorithm.The result shows that the 31 provincial capitals can be divided into five epidemic risk-level regions.The higher risk level,the higher proportion of vaccine coverage among the population to be vaccinated.Aside from that,in terms of reducing the cumulative number of infections,the allocation strategy recommended in this paper is superior to the allocation strategy based on regional population ratio and the preferential full-coverage strategy in high-risk levels.