In the last years, researchers and energy utilities are showing a rising interest in the study and definition of actual buildings’ energy uses. A key aspect of this investigation is the description of daily energy use patterns and their variability over the time. This paper discusses the application of machine learning techniques for pattern recognition with the implementation of a Self-Organizing Map (SOM) algorithm coupled with a k-means clustering algorithm on a dataset of registered electrical energy use in a residential building located in Milan. In the study, five clusters emerged with different daily patterns, that can be ascribed to different uses of electric appliances by people inside the flats.

Pattern Recognition And Classification For Electrical Energy Use In Residential Buildings

Ferrando, Martina;Erba, Silvia;Causone, Francesco;
2020-01-01

Abstract

In the last years, researchers and energy utilities are showing a rising interest in the study and definition of actual buildings’ energy uses. A key aspect of this investigation is the description of daily energy use patterns and their variability over the time. This paper discusses the application of machine learning techniques for pattern recognition with the implementation of a Self-Organizing Map (SOM) algorithm coupled with a k-means clustering algorithm on a dataset of registered electrical energy use in a residential building located in Milan. In the study, five clusters emerged with different daily patterns, that can be ascribed to different uses of electric appliances by people inside the flats.
2020
Proceedings of the 16th IBPSA Conference
9781775052012
File in questo prodotto:
File Dimensione Formato  
BS2019_210750.pdf

accesso aperto

Descrizione: published version
: Publisher’s version
Dimensione 6.28 MB
Formato Adobe PDF
6.28 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11311/1152231
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 5
  • ???jsp.display-item.citation.isi??? 0
social impact