Data-driven model prediction and optimal control for interventional policy of a class of susceptible-infectious-removed dynamics with COVID-19 data

Abstract

Adaptive optimal-control and model prediction are proposed for a class of susceptible-infectious-removed dynamics according to the COVID-19 data. From the practical point of view, data sets of COVID-19 pandemics are daily collected and presented in a discrete-time sequence. Therefore, the discrete-time mathematical model of COVID-19 pandemics is formulated in this work. By developing the time-varying transmission rate, the model's accuracy is significantly contributed to the actual data of the COVID-19 pandemic. Furthermore, the interventional policy is derived by the proposed optimal controller when the closed-loop performance is guaranteed by theoretical aspects and numerical results.