Safety-Critical Control of Compartmental Epidemiological Models with Measurement Delays

TL;DR

This paper presents a control-theoretic approach to ensure safety in epidemiological models by designing controllers that account for measurement delays, effectively bounding infections, hospitalizations, and deaths during COVID-19 spread.

Contribution

It introduces a safety-critical control methodology for compartmental epidemiological models, including delay compensation, to guarantee safe intervention strategies.

Findings

Controllers successfully bound infection numbers under safe limits.

Delay compensation improves intervention effectiveness.

Method applied to COVID-19 data in the USA.

Abstract

We introduce a methodology to guarantee safety against the spread of infectious diseases by viewing epidemiological models as control systems and by considering human interventions (such as quarantining or social distancing) as control input. We consider a generalized compartmental model that represents the form of the most popular epidemiological models and we design safety-critical controllers that formally guarantee safe evolution with respect to keeping certain populations of interest under prescribed safe limits. Furthermore, we discuss how measurement delays originated from incubation period and testing delays affect safety and how delays can be compensated via predictor feedback. We demonstrate our results by synthesizing active intervention policies that bound the number of infections, hospitalizations and deaths for epidemiological models capturing the spread of COVID-19 in the USA.