PREVISÃO DE DADOS DE ALTA FREQUÊNCIA PARA CARGA ELÉTRICA USANDO HOLT-WINTERS COM DOIS CICLOS / FORECASTING HIGH FREQUENCY LOAD DATA USING A DOUBLE CYCLE HOLT WINTERS APPROACH
AUTOR(ES)
CRISTINA VIDIGAL CABRAL DE MIRANDA
DATA DE PUBLICAÇÃO
2007
RESUMO
The forecast for high frequency data is fundamental for the safety and reliability of the electric power system operation. The methods of exponential smoothing, particularly the Holt-Winters approach and its variations, are appropriate for this type of data due to its highly adaptability and robustness. This work seeks to produce forecasts, every fifteen minutes, for a time horizon of seven days ahead, that is, 672 steps of fifteen minutes ahead, for a brazilian load series of an important distribution utility located in the southeast region of the country. It is considered the two-cycle version of the Holt-Winters approach as proposed recently by J. Taylor, including some improvements, such as, the treatment of bank holidays and the influence of temperature, acting as exogenous inputs.
ASSUNTO(S)
temperature high frequency temperatura altas frequencias
ACESSO AO ARTIGO
Documentos Relacionados
- FORECASTING DAILY LOAD DATA USING STRUCTURAL MODELS AND CUBIC SPLINE
- MODELO DE PREVISÃO HORÁRIA DE CARGA ELÉTRICA PARA LIGHT
- Algoritmos de negociação com dados de alta frequência
- Analysis of breaches and co-breaks in financial series: a non-parametric approach data using high frequency.
- LOAD FORECASTING IN POWER SYSTEMS