Please use this identifier to cite or link to this item:
Title: Aplicación de técnicas de aprendizaje automático para la desagregación del consumo energético de viviendas inteligentes
Author: Guasp Alburquerque, Lucía
Director: Monzon Baeza, Victor  
Tutor: Ortega Redondo, Juan Antonio
Abstract: Domestic energy consumption, in most cases, is unknown to users. Consumers cannot access a report indicating how much each device consumes. If this information were accessible, energy efficiency could be improved, thus generating a reduction in the price of electricity bills and pollution, through a reduction in the carbon footprint. Energy disaggregation is a problem of great importance today, closely related to home automation and the transformation of homes towards smart homes. This task must be addressed in a non-invasive way, which means that the consumption of the different devices must be obtained without having to place sensors in each circuit, since this would be inconvenient for the consumers themselves, as well as an additional expense. One way to address this problem in a non-invasive way is to apply artificial intelligence algorithms on a data set of household energy consumption. This TFM focuses on making a comparison on different machine learning models, with the aim of determining an optimal model for energy disaggregation in smart homes. After implementing different algorithms and configurations, the performance of four models was compared: decision tree, random forest, and two neural network models. The results obtained show that random forest and neural networks predict the consumption of the devices in an adequate way, so they could be used in smart homes to improve energy efficiency.
Keywords: machine learning
disaggregation of energy consumption
smart homes
energy optimization
Document type: info:eu-repo/semantics/masterThesis
Issue Date: 31-May-2021
Publication license:  
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
lguaspTFM0521memoria.pdfMemoria del TFM2,24 MBAdobe PDFThumbnail