Power Quality Improvement of Photovoltaic Distributed Generation System using Artificial Neural Network for Environmental Preservation

Most of the economies are emerging with the growth in the renewable energy system. A solar photovoltaic system is one of the good sources of energy among them which provides clean and green energy. As it adds less pollution to the environment and hence advancement in technology of renewable energy system adds great effect on the environmental preservation. This paper describes a 1-Phi photovoltaic distributed generation system having enhanced power quality features. Initially, the system has been implemented by using pulse width modulation-based switching schemes for the smooth control of the power flow between photovoltaic system, grid, and non-linear load. The system involves nonlinear current compensation and capacitor voltage balancing along with maximum power point tracking. Using this model, sample data has been collected for the training and testing of artificial neural networks. The artificial neural network was trained using the scaled conjugate gradient approach. The response of the neural network provides an estimated reference current for the controller to enhance power quality features. The inverter used in this work also acts as a shunt active power filter during night time. The system's result is simulated and validated through MATLAB/Simulink.