Intelligent assistant for electricity management
Autor(a) principal: | |
---|---|
Data de Publicação: | 2022 |
Tipo de documento: | Dissertação |
Idioma: | eng |
Título da fonte: | Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
Texto Completo: | http://hdl.handle.net/10773/38929 |
Resumo: | The steady increase in energy consumption, caused by population growth worldwide, made energy forecasting to be a great tool for governmental and business decision making, such as electricity production or fuel trade. On one hand, accurately following the energy trend to predict how much electricity is expected to be used prevents wasteful spending. On the other hand, being able to detect early signs of equipment malfunction will also prevent additional expenditure. Therefore, creating an agent that’s both capable of predicting the energy to be used soon as well as detect anomalies would be highly helpful. This work describes the development of an intelligent agent to forecast an electricity data time series and use these predictions to detect anomalies in current time. The development focused on using various architectures featuring popular models for time series forecasting, such as LSTM and GRU, attention mechanisms and even using meta-learning models to create the best performing model. With these architectures, it was possible to create a model highly capable of forecasting the complex dataset of over 50 features, getting an average of 0.04 in normalized RMSE in the most relevant variables, and detect anomalies with high precision and reasonable recall, getting a value of 97.3% of precision for the same variables. |
id |
RCAP_4f62a4e7aad53a97edf2071ef6e27539 |
---|---|
oai_identifier_str |
oai:ria.ua.pt:10773/38929 |
network_acronym_str |
RCAP |
network_name_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository_id_str |
7160 |
spelling |
Intelligent assistant for electricity managementAnomaly detectionDeep learningMachine learningMeta-learningTime series forecastingThe steady increase in energy consumption, caused by population growth worldwide, made energy forecasting to be a great tool for governmental and business decision making, such as electricity production or fuel trade. On one hand, accurately following the energy trend to predict how much electricity is expected to be used prevents wasteful spending. On the other hand, being able to detect early signs of equipment malfunction will also prevent additional expenditure. Therefore, creating an agent that’s both capable of predicting the energy to be used soon as well as detect anomalies would be highly helpful. This work describes the development of an intelligent agent to forecast an electricity data time series and use these predictions to detect anomalies in current time. The development focused on using various architectures featuring popular models for time series forecasting, such as LSTM and GRU, attention mechanisms and even using meta-learning models to create the best performing model. With these architectures, it was possible to create a model highly capable of forecasting the complex dataset of over 50 features, getting an average of 0.04 in normalized RMSE in the most relevant variables, and detect anomalies with high precision and reasonable recall, getting a value of 97.3% of precision for the same variables.O aumento de consumo de energia, causado pelo crescimento da população a nível mundial, tornou a previsão de eletricidade uma ferramenta bastante importante para governos e empresas conseguirem fazer decisões importantes, como a produção de eletricidade necessária e comércio de combustivel. Por um lado, seguir a tendência do consumo para prever um valor de eletricidade previne gastos desnecessários. Por outro lado, detetar sinais precoces de anomalias em máquinas pode prevenir gastos adicionais. Portanto, criar um agente que seja tanto capaz de prever os gastos de eletricidade num futuro próximo e capaz de detetar anomalias irá ajudar bastante as empresas que passam por este tipo de decisões. Este projeto descreve o desenvolvimento de um agente inteligente para prever uma série temporal de eletricidade e usar estas previsões para detetar anomalias no tempo atual. O desenvolvimento foca-se em usar várias arquiteturas com modelos populares para previsão de séries temporais, como LSTM e GRU, mecanismos de atenção e até usar modelos meta-learning para criar o modelo com melhor performance. Com estas arquiteturas, foi possível criar um modelo que consiga prever um dataset complexo com mais de 50 features, obtendo, em média, um RMSE normalizado de 0.04 para as variaveis mais relevantes, e detetar anomalias no mesmo com alta precisão, com um valor de 97.3% de precisão para as mesmas variáveis.2023-07-24T09:06:55Z2022-12-21T00:00:00Z2022-12-21info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisapplication/pdfhttp://hdl.handle.net/10773/38929engOliveira, Pedro Miguel Rochainfo: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:RCAAP2024-02-22T12:15:49Zoai:ria.ua.pt:10773/38929Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-20T03:09:08.413777Repositório 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 |
Intelligent assistant for electricity management |
title |
Intelligent assistant for electricity management |
spellingShingle |
Intelligent assistant for electricity management Oliveira, Pedro Miguel Rocha Anomaly detection Deep learning Machine learning Meta-learning Time series forecasting |
title_short |
Intelligent assistant for electricity management |
title_full |
Intelligent assistant for electricity management |
title_fullStr |
Intelligent assistant for electricity management |
title_full_unstemmed |
Intelligent assistant for electricity management |
title_sort |
Intelligent assistant for electricity management |
author |
Oliveira, Pedro Miguel Rocha |
author_facet |
Oliveira, Pedro Miguel Rocha |
author_role |
author |
dc.contributor.author.fl_str_mv |
Oliveira, Pedro Miguel Rocha |
dc.subject.por.fl_str_mv |
Anomaly detection Deep learning Machine learning Meta-learning Time series forecasting |
topic |
Anomaly detection Deep learning Machine learning Meta-learning Time series forecasting |
description |
The steady increase in energy consumption, caused by population growth worldwide, made energy forecasting to be a great tool for governmental and business decision making, such as electricity production or fuel trade. On one hand, accurately following the energy trend to predict how much electricity is expected to be used prevents wasteful spending. On the other hand, being able to detect early signs of equipment malfunction will also prevent additional expenditure. Therefore, creating an agent that’s both capable of predicting the energy to be used soon as well as detect anomalies would be highly helpful. This work describes the development of an intelligent agent to forecast an electricity data time series and use these predictions to detect anomalies in current time. The development focused on using various architectures featuring popular models for time series forecasting, such as LSTM and GRU, attention mechanisms and even using meta-learning models to create the best performing model. With these architectures, it was possible to create a model highly capable of forecasting the complex dataset of over 50 features, getting an average of 0.04 in normalized RMSE in the most relevant variables, and detect anomalies with high precision and reasonable recall, getting a value of 97.3% of precision for the same variables. |
publishDate |
2022 |
dc.date.none.fl_str_mv |
2022-12-21T00:00:00Z 2022-12-21 2023-07-24T09:06:55Z |
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/10773/38929 |
url |
http://hdl.handle.net/10773/38929 |
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_ |
1799137742875000832 |