Multicell Converter Based Adaptive Shunt Active Filter Using Ann to Compensate Var and Nonlinear Load Distortion

Highly distorted non-linear loads such as PV cells, EV chargers, power electronic converters, etc. cause a range of power quality issues including voltage fluctuations, voltage unbalance and harmonics in the distribution networks. While there are a decent number of developed methods and techniques for the purpose of harmonics mitigation and reactive power compensation in distribution networks, this research utilizes the concept of shunt active power filtering (SAPF). SAPFs are generally considered as a practical and effective alternative to classical methods for compensating harmonics and reactive power requirements of nonlinear loads. The SAPF employed in this research work specifically utilizes a double-flying capacitor multicell (DFCM) converter to generate a.c. reference signals and inject them into the network. The implemented converter structure (DFCM) to compensate for network harmonics and reactive power, presents several key advantages over traditional power quality improvement solutions. To obtain the individual harmonic components of the load nonlinear current and then inject proper compensation signals, the research adopts an estimation method based on Adaptive Neural Network (ANN) which in turn offers higher accuracy and fast response in current harmonics identification. To validate the proposed configuration and control strategy, extensive simulations are conducted in IEEE 13 -Bus with a non-linear load in an unbalanced condition using MATLAB Simulink. The simulation results demonstrate the high efficiency and accuracy of the developed method and configuration.