NEURAL NETWORK APPLIED ON A SOLAR ENERGY SYSTEM

Abstract

In this context, the optimization of photovoltaic energy production is studied and presented. In order to extract the maximum available power delivered by the photovoltaic generators (GPV), the technique using an adaptation stage between the generator and the load is chosen. This stage acts as an interface between the two elements by ensuring, through a control action, the transfer of the maximum power supplied so that it is as close as possible to the maximum power. For this purpose, we have been particularly interested in the application of the algorithm based on the neural network in the control of this adaptation stage associated with the photovoltaic generator to ensure its operation at its point of maximum power. Simulation results show that this system adapts to variations in external disturbances and has a relatively small peak at transient state and a uniform steady-state speed. The controller gives satisfactory results and shows their effectiveness not only in tracking the point of maximum power but also in response time.