Harmonics Reduction on Electric Power Grid Using Shunt Hybrid Active Power Filter with Finite-Control-Set Model-Predictive Control

Abstract

This paper discusses a technique of reducing harmonics in electric power grids using Shunt Hybrid Active Power Filter. The filter is controlled by a Model-Predictive Control technique. It can predict the output current magnitude based on its predictive model by noticing the constructed grid topology. Then, this current prediction has been compared to reference currents in order to get an optimal switching pattern concerning a cost function while the optimization has been conducted in order to figure out the maximum condition for the switching pattern. In addition, the Shunt Hybrid Active Power Filter has been compared by combining passive and active filters, which has been expected to be able to reduce harmonics optimally. The use of an LC filter in a single tuned has been also targetted in order to reduce the harmonics in the fifth-order, while the active filter would reduce harmonics due to their resonances. The constructed model has been tested in the simulation in order to evaluate how much harmonic reduction could be performed by the model, and how this model has coped with an unbalanced load, and later affecting the electric power quality. The simulation shows that Shunt Hybrid Active Power Filter along with the Model-Predictive Control method tested by non-linear loads, in terms of balanced and unbalanced loads, can reduce Total Harmonic Distortion of current loads to under 1%. As for unbalanced loads, the phase angle on source voltage does not encounter displacement. Therefore, the Model-Predictive Control on Shunt Hybrid Active Power Filter can effectively be used to reduce harmonics in the grid, during both balanced and unbalanced loading condition.