Volume 1 - No 2
MODEL BASED ON COMBINATION OF PREDICTORS FOR SHORT-TERM LOAD FORECAST
Abstract
The forecasting of load demand is a fundamental task for the proper functioning of electrical
systems, because many decision-making processes such as planning, operation, security analysis
and market decisions are highly influenced by this process. Knowing that meet the demand charge
is a stochastic process, this prediction proves important to a energy company to operate safely and
economically. From this viewpoint, this paper proposes the development of a methodology for load
forecasting in the short term, based on a combination of several different predictors. The results
showed that the combination of the predictors presented in most cases examined, more accurate
results when compared to results obtained for only a forecast component individually. In this
model, the forecast of the load demand curve was based on demand curves for known days.
Paper Details
PaperID: p. 20-31
Author's Name: Max Olinto Moreira, Ricardo Menezes Salgado, Takaaki Ohishi, Eric Batista Ferreira
Volume: Volume 1
Issues: No 2
Keywords: Load Forecasting, Ensemble Model, Artificial Intelligence, Power System Planning
Year: 2011
Month: December
Pages: 20-31