ANFIS-based Power Quality Improvement by Photovoltaic Integrated UPQC at Distribution System

Abstract

The advent of power electronic devices for the control of non-linear loads has made an impact on the power quality at the Distribution side of the utility grid. The Distribution side must be compensated for both reactive and real power compensation while simultaneously improving the power quality. Unified Power Quality Conditioner is a FACTS device with back-to-back converters coupled together with a DC link element that improves the power quality at the Distribution side. The DC link element sizing and the converter ratings are the challenges faced during the design of UPQC. The Distributed Energy Resource (DER) integrated at the DC link element, supports for power exchange and minimization of the converter ratings. UPQC integrated with a DER simultaneously improves the fault-ride-through capabilities at the Point of Common Coupling. The PV integrated UPQC with bidirectional series and shunt converters are controlled by a hybrid combination of Unit Vector Template and p-q theory, respectively. The control unit of the proposed system has to be more precise in a closed-loop structure. The conventional mathematical based PI controllers fail to perform well during the transient oscillations for the UPQC system. The proposed PV-UPQC system has been analysed with an Adaptive Neuro-Fuzzy controller that embeds a reinforced learning algorithm. The Fuzzy-Model-Based (FMB) controller improves the performance of the system by inferencing the system parameters through linguistic rules and helps in reference current generation. The PV-UPQC performed significantly well even during various load conditions. The implementation of an Adaptive Neuro-Fuzzy Inference System leads to minimization in the percentage of Total Harmonic Distortion level.