A Solar Thermal System Temperature Prediction of a Smart Building for Data Recovery and Security Purposes

  1. José-Luis Casteleiro-Roca 1
  2. Maite García-Ordás 2
  3. Esteban Jove Pérez 1
  4. Francisco Zayas-Gato 1
  5. Héctor Quintián Pardo 1
  6. Héctor Alaiz Moretón 2
  7. José Luis Calvo Rolle 1
  1. 1 Universidade da Coruña

    Universidade da Coruña

    La Coruña, España

    ROR https://ror.org/01qckj285

  2. 2 Universidad de León

    Universidad de León

    León, España

    ROR https://ror.org/02tzt0b78

Intelligent Data Engineering and Automated Learning – IDEAL 2020. 21st International Conference: Guimarães, Portugal; November 4–6, 2020. Proceedings
  1. Cesar Analide (ed. lit.)
  2. Paulo Novais (ed. lit.)
  3. David Camacho Fernández (ed. lit.)
  4. Hujun Yin (ed. lit.)

Publisher: Springer International Publishing AG

ISBN: 978-3-030-62362-3

Year of publication: 2020

Volume: 2

Pages: 468-476

Congress: Intelligent Data Engineering and Automated Learning – IDEAL (21. 2020. Guimarães)

Type: Conference paper


This paper perform a comparison between different clustering algorithms, that their optimal number of clusters has been calculated throw different performance measurements. The comparison takes into account the prediction of the thermal solar panel output temperature to conclude what is the best clustering division. The used dataset is extracted from a Bioclimatic house that belongs to Sotavento Galicia Foundation, and it is composed of the most important variables in the thermal solar energy generation system.Silhouette, Calinski-Harabasz, and Davies-Bouldin were used to achieve the optimal number of clusters and then, Artificial Neural Networks and Polynomial Regression were trained, with several configurations, to create a hybrid intelligent model for regression. Very good results were obtained with this procedure, that allows to reduce the computational cost of creating a hybrid model without knowing the number of clusters for the dataset.