Mathematical Modeling for Spread and Control of COVID-19

Abstract

In order to accurately simulate the spread of the coronavirus disease 2019 (COVID-19) and provide certain reference for formulation of epidemic prevention and control policies, a new nonlinear dynamics model is presented. Considering real prevention and control measures, the total population is divided into seven groups in the model, including susceptible, latent, quarantined, asymptomatic, symptomatic, hospitalized and recovered groups. The basic reproduction number is calculated and analyzed. The cure rate and mortality rate are fitted as time-varying functions, and the remaining parameters and initial values of some states are fitted by least squares with the number of currently confirmed cases. The real data of COVID-19 in Wuhan and other regions of Hubei Province within 20 days (February 14 to March 4, 2020) are used to simulate and verify the model. Simulation results show that the mean relative error of fitting is 0.629%. Compared with the real data of cumulatively confirmed cases, cured cases and deaths, the mean relative errors of the epidemic prediction by the model in Wuhan are 0.772%, 3.517%, and 2.025%, respectively, and in other regions of Hubei Province are 0.808%, 2.241% and 2.39%, respectively. It indicates that the model can accurately simulate the spread of COVID-19 and has a wide range of applicability. Moreover, sensitivity analysis is made and simulated on the model parameters. The impacts of various prevention and control measures on the spread of COVID-19 are discussed. The analysis results show that the most effective prevention and control measures include reducing contact among people, strengthening tracking and quarantine of close contacts, and increasing detection and treatment capabilities so that the infected people can be isolated and treated as soon as possible. © 2021, Editorial Office of Journal of Xi'an Jiaotong University. All right reserved.