Adaptive Switching Algorithm for Shunt Active Power filter with Model Predictive Control

Abstract

The Algorithmic controllers have been widely adopted for power converter applications as in Model Predictive control and adaptive control. Algorithmic controllers are usually functioning with restrictions for electronic component protection. This article presents algorithmic control for modulating the switching frequency of power converters to reduce both switching losses and switching frequency. A low volt converter for shunt active power filter is used as an application for showing the effect of adaptive switching. Least Mean Square Adaptive Linear Neuron (ADALINE LMS) is proposed as algorithmic adaptive switching for model predictive controllers used with power converter applications. The proposed technique shows the reliability of limiting the switching frequency, interfacing it with model predictive control. The new technique is applicable on electrical grid and aircraft applications. THD is kept within a stable acceptable limit <5% with less switching frequency. The technique is designed and simulated on MATLAB SIMULINK and results are verified.