Analysis of infection spread using simulation of infection based on population distribution data

Abstract

Considering the spread of COVID-19 infection and the migration to Category V Infectious Diseases, it is important to continuously detect the spread of infection with a small amount of data. We have developed a multi-agent simulation reflecting human behavior based on the power law and small-world networks. We used the number of infected persons to estimate the parameters of the simulation, but it became difficult to use the reported number of infected persons in the simulation due to the shift from total identification of infected persons to fixed-point identification and people's interest in COVID-19 decreased. In this study, we constructed a simulation that can continue to analyze infections even after the total number of infections is no longer available by incorporating mobile spatial data. With the simulation, we reproduced infection and estimate the number of infected persons in the future, and we showed that we can continuously analyze the spread of infection.