| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 2.36 MB | Adobe PDF |
Authors
Abstract(s)
As metodologias de previsão de cargas elétricas têm sido alvo de um desenvolvimento considerável nos últimos anos, sendo hoje parte integrante dos atuais sistemas de planeamento e operação que se encontram ao dispor de empresas operadoras dos sistemas de energia elétrica.
Este desenvolvimento tem como principal causa a reestruturação do setor elétrico que conduziu à liberalização do mesmo em vários países. A liberalização, por sua vez trouxe consigo o aumento da complexidade organizativa e consequentemente a necessidade da previsão do consumo de energia elétrica.
Este trabalho incide sobre a problemática da previsão do consumo de energia elétrica em horizontes temporais de curto prazo. Tendo como ponto de partida os dados do passado relativos aos consumos verificados, disponibilizados pela REN, efetuar-se-á a análise do comportamento desta série temporal e posteriormente a previsão do consumo de energia elétrica.
Este estudo propõe e compara vários modelos de alisamento exponencial de Holt-Winterscom dupla sazonalidade.
Abstract: Load forecast methodologies had a significant development in recent years and play an important role for electric power systems corporations aim to render more efficient procedures like operation and planning of electric energy systems. The reorganization of electric sector that led to the market liberalization in several countries was the main cause of this load forecast methodologies development. This development happened due to increased complexity of involved organizations and consequently the necessity to predict the energy consumption. This work is about the issue of the load forecast in short periods. It attempts that knowing the energy consumes past data provided by REN and analyzing that time series, it will be able to have an electric load forecast. This thesis proposes and compares several double seasonal Holt-Winters exponential smoothing methods.
Abstract: Load forecast methodologies had a significant development in recent years and play an important role for electric power systems corporations aim to render more efficient procedures like operation and planning of electric energy systems. The reorganization of electric sector that led to the market liberalization in several countries was the main cause of this load forecast methodologies development. This development happened due to increased complexity of involved organizations and consequently the necessity to predict the energy consumption. This work is about the issue of the load forecast in short periods. It attempts that knowing the energy consumes past data provided by REN and analyzing that time series, it will be able to have an electric load forecast. This thesis proposes and compares several double seasonal Holt-Winters exponential smoothing methods.
Description
Dissertação para a obtenção do grau de Mestre em Engenharia Electrotécnica - Ramo de Energia
Keywords
Previsão Forecasting Série temporal Time series Energia elétrica Electrical energy Alisamento exponencial Exponential smoothing
Citation
ALVES, Pedro Miguel Marques - Previsão de curto prazo do consumo de energia elétrica. Lisboa: Instituto Superior de Engenharia de Lisboa, 2013. Dissertação de mestrado.
Publisher
Instituto Superior de Engenharia de Lisboa