Indexado en

Licencia y uso

Citaciones

Cited 1 times in

Cited 1 times in

Altmetrics

Grant support

The Universidad de Valladolid supported this study with the predoctoral contracts of 2020, co-funded by Santander Bank. This work has been financed also by the Spanish Ministry of Science and Innovation, under project PID2020-113533RB-C33. The Universidad de Valladolid also supported this study with ERASMUS+ KA-107. Finally, we have to thank the MOVILIDAD DE DOCTORANDOS Y DOCTORANDAS UVa 2023 from the University of Valladolid. We also appreciate the help of other members of our departments.

Análisis de autorías institucional

Compartir

Enhancing Solar Cell Classification Using Mamdani Fuzzy Logic Over Electroluminescence Images: A Comparative Analysis with Machine Learning Methods

Publicado en:Communications in Computer and Information Science. 1938 159-173 - 2024-01-01 1938(), DOI: 10.1007/978-3-031-52517-9_11

Autores: Mateo-Romero, Hector Felipe; Carbono dela Rosa, Mario Eduardo; Hernandez-Callejo, Luis; Gonzalez-Rebollo, Miguel Angel; Cardenoso-Payo, Valentin; Alonso-Gomez, Victor; Gallardo-Saavedra, Sara

Afiliaciones

Resumen

This work presents a Mamdani Fuzzy Logic model capable of classifying solar cells according to their energetic performance. The model has 3 different inputs: The proportion of black pixels, gray pixels, and white pixels. One additional output for informing of possible bad inputs is also provided. The three values are obtained from an Electroluminescence image of the cell. The model has been developed using cells whose performance has been obtained by measuring the Intensity-Voltage Curves of the cells. The performance of the model has been shown by testing it with a validation set, obtaining a 99.0% of accuracy, when other methods such as Ensemble Classifiers and Decision Trees obtain a 97.7%. This shows that the presented model is capable of solving the problem better than traditional Machine Learning methods.

Palabras clave