Application of Time Delay Recurrent Neural Network for Shunt Active Power Filter in 3-Phase Grid-tied PV System

In this paper, time-delay recurrent neural network (TDRNN) has been designed and implemented for the power quality improvement in grid-connected photovoltaic (PV) system. Shunt active power filter (SAPF) is commonly used for mitigation of power quality (PQ) problems. TDRNN algorithm has been used for estimating the fundamental load component, which is used further for the generation of switching pulses for shunt active power filter (SAPF). The proposed algorithm has been justified by simulation and experimental outcomes.