A Renewable Energy Source Fed Neuro-Fuzzy Controlled Multilevel UPQC for Power Quality Improvement

This study examines the H-Bridge cascade five-level unified Power quality conditioner coupled with photovoltaic and battery storage systems to handle the power quality related problems. To eliminate the requirement of the complex transformations like abc, dq0, ?? , the artificial neural network-based control scheme with levenberg- marquardt back propagation training method is adopted for the H-Bridge cascade 5L-UPQC to produce the necessary reference signals for the shunt and series voltage source converters. In addition, an adaptive neuro-fuzzy controller is suggested for DC link error current minimization, which utilizes the properties of both the fuzzy as well as ANN. The prime goal of the developed scheme is to maintain constant DLCV in load changes, diminish of THD in the source current and load voltage, and maximum elimination of supply voltage fluctuations like sag/swell and disturbances. The suggested method was demonstrated in four cases with several permutations of loads. However, to reveal the performance of the developed method, the comparison is carried out with the two level and three level UPQC configurations in addition to methods available in literature.