Design of Robust Controllers Applied to DC-DC Converters with Uncertain Parameters Optimized by Metaheuristics

Abstract

This work presents a design procedure for optimized robust controllers applied to DC-DC boost converters with uncertain parameters. The converter is represented by a polytopic model, to include parametric uncertainties in the resistive load, in the input voltage and in the duty cycle. In order to guide the design of the control gains, an optimization problem is formulated based on the minimization of a cost function, which demands from the designer only the choice of reference values for phase margin and crossover frequency. The controller obtained by the proposed procedure, based on a metaheuristics algorithm of simple computational implementation, is robust and guarantees better responses than a controller optimized only for a nominal model of the converter, obtained by an usual routine in the literature, as illustrated by the results.