Supervisory control of PV-battery systems by online tuned neural networks

The paper deals with a neural network based supervisor control system for a PhotoVoltaic (PV) plant. The aim of the work is to feed the power line with the 24 hours ahead forecast of the PV production. An on-line self-learning prediction algorithm is used to forecast the power production of the PV plant. The learning algorithm is based on a Radial Basis Function (RBF) network and combines the growing criterion and the pruning strategy of the minimal resource allocating network technique. The power feeding the electric line is scheduled by a Fuzzy Logic Supervisor (FLS) which controls the charge and discharge of a battery used as an energy buffer. The proposed solution has been experimentally tested on a 14 KWp PV plant and a lithium battery pack.