Hybrid-MPC applied to epidemiology

The classic Susceptible-Infected-Removed (SIR) model provides a basis for understanding the propagation of infectious diseases through a set of differential equations. The standard SIR model compartmentalizes the overall population into three classes: the Susceptible, the Infected, and the Recovered. A susceptible person when exposed to an infected person becomes infected themselves. However, no transfer takes place in the reverse direction. This process draws an analogy from an irreversible auto-catalytic reaction. The removal happens from the infected population through either recovery or death.

Categorical Control of SIR Infection using 3DoF Hybrid MPC

Mixed Logic Dynamic Modeling of SIR Problem

3DoF applied to SIR model

• Measured disturbance (Δkv = 0.25 Day−1) at t = 80 Days.

• Unmeasured disturbance (Δγ = 0.1 Day−1) at t = 120 Days.

• Reconfiguration bounds on the population = [550 900].

• No Chattering. Slack enforces soft constraints.

• Unavoidable yet Acceptable offset.