Artificial Intelligence Applied to the Thermal Characterization of Building Integrated Photovoltaic Technologies

Abstract

Building Integrated Photovoltaic (BIPV) systems aim not only to generate part of the electricity consumed by the edifice, but also to reduce the environmental impact, such as the Green House Gases emissions produced by the generation of electricity by fossil fuels and the land area that would be required when installed the PV devices in floor. Nevertheless, the thermal behaviour of the PV modules has an impact into the indoor temperature of the building. To study the thermal effect produced by the environmental factors into the module temperature, Artificial intelligence, i.e. Genetic Programming (GP), is applied to one year of data of crystalline silicone PV modules mounted as building elements at outdoor operating conditions to get the model which best describes the thermal behaviour. Two well-known models guided the algorithm, the ones proposed by Ross and by Sandia Laboratories. From the application of Genetic Programming, a model was obtained which calculates the module temperature with less than 1% error. Six constants parameters are obtained for the thermal behaviour of crystalline silicon PV technology, fitting an equation whose structure reflects the two well-known models of Ross and Sandia.

Lucía Serrano-Luján

Associate Professor

Lucía Serrano-Luján is an Associate Professor in the Department of Computer Science. Her research field is multidisciplinary. She developed a Life Cycle Assessment methodology to assess renewable energies and applied AI to their data. Her main goal is to impact energy-related materials production and find a more sustainable way to develop them. She has applied LCA to reduce graphene oxide and perovskites solar cells, build integrated photovoltaics, etc.

J. Manuel Colmenar

Full Professor

My research interests are focused on metaheuristics applied to optimization problems. I have worked on different combinatorial optimization problems applying trajectorial algorithms such us GRASP or VNS. Besides, I am very interested in applications of Grammatical Evolution, specifically in model and prediction domain, as alternative to machine learning approaches.