-
Fabiane Souza
Universidade Federal de Lavras, MG
-
Sílvia C. Ferreira
Universidade Federal de Lavras, MG
-
Robson B. Gonzatti
Universidade Federal de Itajubá, MG
-
Rondineli R. Pereira
Universidade Federal de Itajubá, MG
-
João Gabriel L. Foster
Universidade Federal de Itajubá, MG
-
Joselino S. Filho
EDP Brasil, SP
Keywords:
Parameter Mismatch, Adaptive Notch Filter, Least Mean Square
Abstract
Hybrid Active Power Filters (HAPF) have attracted great interest due to their ability to compensate for harmonics and reactive power. For reactive compensation purposes, the Finite Control Set Model Predictive Control (FCS-MPC) has shown to be very efficient in the HAPF used in this work. To accomplish this task, a mathematical model of the equipment is deduced aiming to perform prediction steps and calculate the control references. The parameters of this model are dependent on the values of its passive components. However, these components may change due to failure conditions or time degradation. The proposal of this paper is to analyze the influence of the variation of the HAPF parameters on the voltage and currents used as control references. As a contribution, it is proposed to estimate these parameters using an Adaptive Notch Filter with Least Mean Square algorithm. Simulations results with MATLAB/Simulink show that there is a high influence of capacitor bank impedance mismatch in the FCS-MPC and that the proposed algorithm is able to estimate this parameter properly.