A Solar Module Power Production Model Predictor for Automated Solar Module Manufacturing

Corrado Santoro, Federico Fausto Santoro, Eleonora Arena, Andrea Canino

Abstract: Sunlight is one of the most abundant and freely available energy resources on our planet. The capacity to transform this solar energy in electrical power depends on the quality and materials of solar photovoltaic modules. The production process of cells and modules consists of a chain of physical and chemical steps. A solar module is a set of solar cells, that are the core of energy conversion technology interconnected together. These are classified considering their efficiency in terms of the maximum power that a cell can produce with a specific calibrated light exposition. Considering these factors this paper proposes a neural network model, trained with a real dataset, capable of predicting the final maximum power of solar modules by using a set of solar cells as input. Furthermore, the proposed model works also with more than one class of solar cells. Finally, the accuracy of prediction has reached an acceptable value and a final software has been developed to perform these predictions.