Performance of adaptive radial basis functional neural network for inverter control

A single phase grid connected photovoltaic (PV) system is susceptible to a number of power quality (PQ) problems, including power factor, harmonic current, voltage fluctuations, and load unbalance. Compensation is needed to address these PQ concerns. In this study, a single phase shunt active power filter is presented to handle power quality issues using novel and straight forward radial basis function neural network (RBFNN) controller architecture and to ensure maximum power flow between PV and grid using a maximum power point tracker control technique. The design takes into account a single neuron in the hidden layer, and the network is trained on-line to be suitable for inverter control to reduce power quality (PQ) issues. The newly developed controller has a single input for the load current and is able to isolate the fundamental component of the current. Tracking is fast and achieved within one cycle. The trained model shows exceptional results for load compensation under various loading conditions. With the suggested RBFNN controller, both findings from simulation and from experiments have been shown to work.