Controle de Tensão em Conversores Boost com FCS-MPC - um Estudo Comparativo de Diferentes Funções Custo

Abstract

The Finite Control Set Model Predictive Control (FCS-MPC) when applied for voltage control in a Boost converter presents limitations due to the non-minimum phase characteristics of this system (NMP). Therefore, setting an adequate cost function is essential for the success of the control strategy in short-time prediction horizons. In this work, the performance of the FCS- MPC applied to the Boost converter is compared among different cost-functions configurations. In a first scenario, the traditional voltage cost function is compared with two distinct proposals made by the authors: a multivariable voltage and current cost function, and a current-based function that perform indirect voltage control. In a second scenario, model corrections are made as proposed by Villarroel et al. (2019) for the delays caused by the non-minimum phase. The converter is software-simulated using Matlab/Simulink and the results obtained shows that the use of the multivariable cost function alone does not solve the optimization problem caused by the non-minimum phase characteristic. However, when the system is indirectly controlled using current control, or else when is operated with FNM corrections, the FCS-MPC demonstrated to be able to optimize the system output with low steady-state errors. Furthermore, the use of the multivariable cost function with corrections slightly improves the output voltage oscillations while maintaining a very fast transient response characteristic.