Discrete-time switching MPC with applications to mitigate resistance in viral infections

Many engineering applications can be described as switched linear systems, in which the manipulated control action is the time-dependent switching signal. In such a case, the control strategy must select a linear autonomous system at each time step, among a finite number of them. Even when this selection can be done by solving a Dynamic Programming (DP) problem, the implementation of such a solution is often difficult and state/control constraints cannot be explicitly accounted for. In this paper, a new set-based Model Predictive Control (MPC) strategy is presented to handle switched linear systems in a tractable form. The optimization problem at the core of the MPC formulation consists of an easy-to-solve mixed-integer optimization problem, whose solution is applied in a receding horizon way. The medical application of viral mutation and its respective drug resistance is addressed to acute and chronic infections. The objective is to attenuate the effect of mutations on the total viral load, and the numerical results suggested that the proposed strategy outperforms the schedule for available treatments.