Short-term forecasting for household electricity load with dynamic feature selection using power cepstrum
Autor(a) principal: | |
---|---|
Data de Publicação: | 2023 |
Tipo de documento: | Dissertação |
Idioma: | eng |
Título da fonte: | Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
Texto Completo: | http://hdl.handle.net/10362/149105 |
Resumo: | Dissertation presented as the partial requirement for obtaining a Master's degree in Data Science and Advanced Analytics, specialization in Data Science |
id |
RCAP_117462973fe7a0627646ec135f54d1f9 |
---|---|
oai_identifier_str |
oai:run.unl.pt:10362/149105 |
network_acronym_str |
RCAP |
network_name_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository_id_str |
7160 |
spelling |
Short-term forecasting for household electricity load with dynamic feature selection using power cepstrumShort-term energy load ForecastingDynamic selectionPower CepstrumMachine LearningConvLSTMDomínio/Área Científica::Ciências Naturais::Ciências da Computação e da InformaçãoDissertation presented as the partial requirement for obtaining a Master's degree in Data Science and Advanced Analytics, specialization in Data ScienceElectrical energy is present in our civilization and has positive and negative impacts in our environment, renewable and green energy like solar and wind energy works with significant less negative environmental impacts, reducing disasters and fuel dependency. Although, the transition to renewable and green energy demands advanced technologies to manage energy distribution in society, since the clean sources are stochastic. In this study the research will be done to improve electricity consumption forecasting in a household, a tool that can help the energy distribution management by microgrids to determine the amount of energy used by consumers at a particular moment, resulting in reducing energy waste and allowing P2P energy trading. The goal of this study is to do short-term forecasting and test the ability of Power Cepstrum to select autoregressive features, the dataset used is of minute-by-minute electricity consumption in kilowatts of a single household in the town of Sceaux,, France, between December 2006 and November 2010, the model tested was the Convolutional Long Short-Term Memory neural network with selected auto-regressive feature model (CLSAF), a Convolutional Long Short-Term Memory model working with Persistence model with dynamic feature selection, the ability of Power Cepstrum to select the autoregressive feature is tested and compared to CLSAF using Autocorrelation Function to select the autoregressive feature, the results are compared either to state of art models such as ConvLSTM and Persistence model. The tests were done comparing different theta threshold, input lags, resolutions, and input length. The result show that Power Cepstrum can be used as a replacement for Autocorrelation Function, CLSAF have comparable accuracy to ConvLSTM model and better runtime performance when using y[t-1] as input lag, for 30 minutes resolution is possible to observe great difference between runtime prediction without losing accuracy performance, Power Cepstrum showed better runtime prediction when compared to autocorrelation function, also, higher input length improved models performance.Scott, Ian JamesRUNAgottani, Luis Fernando Rodrigues2023-02-13T13:24:21Z2023-01-242023-01-24T00:00:00Zinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisapplication/pdfhttp://hdl.handle.net/10362/149105TID:203221338enginfo:eu-repo/semantics/openAccessreponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãoinstacron:RCAAP2024-03-11T05:30:51Zoai:run.unl.pt:10362/149105Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-20T03:53:36.822165Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãofalse |
dc.title.none.fl_str_mv |
Short-term forecasting for household electricity load with dynamic feature selection using power cepstrum |
title |
Short-term forecasting for household electricity load with dynamic feature selection using power cepstrum |
spellingShingle |
Short-term forecasting for household electricity load with dynamic feature selection using power cepstrum Agottani, Luis Fernando Rodrigues Short-term energy load Forecasting Dynamic selection Power Cepstrum Machine Learning ConvLSTM Domínio/Área Científica::Ciências Naturais::Ciências da Computação e da Informação |
title_short |
Short-term forecasting for household electricity load with dynamic feature selection using power cepstrum |
title_full |
Short-term forecasting for household electricity load with dynamic feature selection using power cepstrum |
title_fullStr |
Short-term forecasting for household electricity load with dynamic feature selection using power cepstrum |
title_full_unstemmed |
Short-term forecasting for household electricity load with dynamic feature selection using power cepstrum |
title_sort |
Short-term forecasting for household electricity load with dynamic feature selection using power cepstrum |
author |
Agottani, Luis Fernando Rodrigues |
author_facet |
Agottani, Luis Fernando Rodrigues |
author_role |
author |
dc.contributor.none.fl_str_mv |
Scott, Ian James RUN |
dc.contributor.author.fl_str_mv |
Agottani, Luis Fernando Rodrigues |
dc.subject.por.fl_str_mv |
Short-term energy load Forecasting Dynamic selection Power Cepstrum Machine Learning ConvLSTM Domínio/Área Científica::Ciências Naturais::Ciências da Computação e da Informação |
topic |
Short-term energy load Forecasting Dynamic selection Power Cepstrum Machine Learning ConvLSTM Domínio/Área Científica::Ciências Naturais::Ciências da Computação e da Informação |
description |
Dissertation presented as the partial requirement for obtaining a Master's degree in Data Science and Advanced Analytics, specialization in Data Science |
publishDate |
2023 |
dc.date.none.fl_str_mv |
2023-02-13T13:24:21Z 2023-01-24 2023-01-24T00:00:00Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/masterThesis |
format |
masterThesis |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://hdl.handle.net/10362/149105 TID:203221338 |
url |
http://hdl.handle.net/10362/149105 |
identifier_str_mv |
TID:203221338 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
application/pdf |
dc.source.none.fl_str_mv |
reponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação instacron:RCAAP |
instname_str |
Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
instacron_str |
RCAAP |
institution |
RCAAP |
reponame_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
collection |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository.name.fl_str_mv |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
repository.mail.fl_str_mv |
|
_version_ |
1799138126585659392 |