An Artificial Neural Network Based Power Converter for the Control of Positive and Negative Sequence Currents in Unbalance Ac Source

Abstract

The switching strategy for a AC/DC or DC/AC converter by Neural Networks is proposed, with this application the power controllability can be easily controlled by selecting proper positive and negative sequence currents. With the advert of utilising zero sequence components the enhancement of power control by quick switching response, easy structure and better regulated output. This proposed system is designed to avoid over current problems in unbalanced ac sources caused by generator/grid faults. Additionally it avoids the required extra power loop or hysteresis controller which simplifies the control algorithm. The validity of the developed control system is investigated via computer simulations using MATLAB/Simulink.