Optimal Control Concerns Regarding the COVID-19 (SARS-CoV-2) Pandemic in Bahia and Santa Catarina, Brazil

TL;DR

This paper explores the application of Model Predictive Control (MPC) to design social distancing policies for COVID-19 in two Brazilian states, using SIRD models to optimize pandemic mitigation strategies.

Contribution

It introduces a novel MPC framework tailored for pandemic control, comparing centralized and decentralized strategies for different regional contexts.

Findings

Decentralized MPC outperforms centralized in certain scenarios.

Optimized social distancing policies effectively reduce infection peaks.

Framework provides guidelines for future pandemic response planning.

Abstract

The COVID-19 pandemic is the profoundest health crisis of the 21rst century. The SARS-CoV-2 virus arrived in Brazil around March, 2020 and its social and economical backlashes are catastrophic. In this paper, it is investigated how Model Predictive Control (MPC) could be used to plan appropriate social distancing policies to mitigate the pandemic effects in Bahia and Santa Catarina, two states of different regions, culture, and population demography in Brazil. In addition, the parameters of Susceptible-Infected-Recovered-Deceased (SIRD) models for these two states are identified using an optimization procedure. The control input to the process is a social isolation guideline passed to the population. Two MPC strategies are designed: a) a centralized MPC, which coordinates a single control policy for both states; and b) a decentralized strategy, for which one optimization is solved for each state. Simulation results are shown to illustrate and compare both control strategies. The framework serves as guidelines to deals with such pandemic phenomena.