MODEL-FREE ADAPTIVE CONTROL WITH TUNED PARAMETERS BY MULTI-OBJECTIVE DIFFERENTIAL EVOLUTION

JUDAS TADEU GOMES DE SOUSA¹; ALUIZIO FAUSTO RIBEIRO ARAUJO²

¹ UNIVERSIDADE FEDERAL DO VALE DO SÃO FRANCISCO; ² UNIVERSIDADE FEDERAL DE PERNAMBUCO

Baixar PDF doi:10.20906/CPS/CBA2018-0608

Resumo

In recent decades, the development of model free control techniques, whose control is based on measured data during the execution of the process, has intensified. In this context, the data-driven control technique called Model-Free Adaptive Control has highlighted because its applications include linear and nonlinear systems, SISO or MIMO. In this paper, we dealt with an open question in this technique, which is the choice of controller parameters that provide better performance to the control system. To solve this problem, we used a metaheuristic based in a multi-objective differential evolution algorithm to tune the controller pa-rameters. The results confirmed the validity of the proposed strategy, to define the controller parameters, for examples already presented in the literature.

Palavras-chave: Data-driven control; Model-free adaptive control; Nonlinear systems; Multi-objective optimization; Differential evolution