An EKF prediction of COVID-19 propagation under vaccinations and viral variants

Abstract

The COVID-19 pandemic continues to pose significant challenges to global public health, requiring advanced predictive mathematical models for prediction, prevention and control. This paper proposes a novel approach to dynamic estimation of COVID-19 pandemic in the presence of vaccinations and viral variants. By introducing the vaccinated compartment and re-infection factor into the classical susceptible, exposed, infectious, recovered, and deceased (SEIRD) model to characterise the vaccination and re-infection effects, a new vaccination-SEIRD (V-SEIRD) model is established to depict the dynamics of COVID-19 transmission in the presence of vaccinations and viral variants under the variable total population. Upon this model, an extended Kalman filter (EKF) is further developed to simultaneously estimate the model parameters and predict the transmission state for COVID-19 pandemic. Results demonstrate that the suggested approach is capable of characterising the vaccination and re-infection impacts on COVID-19 evolution, resulting in enhanced accuracy for COVID-19 prediction in the presence of vaccinations and viral variants. The proposed method can aid the design of vaccination strategies and public health policies for infectious disease prevention and control.