Vaccine Distribution Modelling in Pandemics through Multi-Agent Systems: COVID-19 Case

Multi-agent systems (MAS) have become increasingly used in addressing complex problems that require collaboration and coordination between multiple entities. In the context of infectious disease management, MAS can enable a more nuanced understanding of the complex social dynamics at play, including the behavior and interactions of individuals that determine the rate of infection and mortality.This paper proposes two vaccine distribution methods based on MAS and the concepts of bankruptcy and the Susceptible-Exposed-Infectious-Recovered (SEIR) model, which take into account the prioritization of individuals based on their behavior and their impact on the infection rate of the disease. To propose these methods first, we propose a base algorithm named Vaccine Allocation Problem (VAP) that prioritizes individuals based on their impact on community health.The proposed methods were evaluated in comparison with classic bankruptcy methods such as Proportional (P), Talmud, and Pinile methods. In addition, we use the Covid-19 data set provided by "Belgian Institute for Health" as a case study. Simulation results indicated the better performance of the proposed method in terms of infection, hospitalization, Intensive Care Unit (ICU), new-In, and mortality rates.

I. Introduction

The COVID-19 pandemic, caused by the coronavirus, emerged in December 2019 and has since spread rapidly worldwide. The World Health Organization (WHO) declared it as a pandemic in March 2020 [6], leading to global efforts to diagnose, prevent, and treat the disease. The recent development of COVID-19 vaccines has been a major breakthrough in combating the virus [12], with reports indicating their effectiveness in preventing the disease and reducing the spread of new mutations [21].