Power Demand Forecast Using Least-Squares Support Vector Machines

Author

dos Santos Coelho, L. ; Klein, C.E.

Author_Institution

Ind. & Syst. Eng. Grad. Program, Pontifical Catholic Univ. of Parana, Curitiba, Brazil

fYear

2009

fDate

8-12 Nov. 2009

Firstpage

1

Lastpage

5

Abstract

This paper aims to share the results on forecasting power demand using least-squares support vector machines. The development is based on model estimation taking in consideration the past measurements for power demand and ambient temperature. All approximated models were evaluated using the multiple correlation coefficient (R²) or mean absolute percentage error (MAPE) and maximum error combined as quality parameters.

Keywords

demand forecasting; least squares approximations; load forecasting; power engineering computing; power system planning; support vector machines; ambient temperature; least squares support vector machines; maximum error; mean absolute percentage error; model estimation; multiple correlation coefficient; power demand forecast; quality parameter; Demand forecasting; Power demand; Power engineering and energy; Power generation; Power measurement; Power system modeling; Support vector machines; System identification; Systems engineering and theory; Temperature; Power Forecasting; Support Vector Machine; System Identification;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent System Applications to Power Systems, 2009. ISAP '09. 15th International Conference on

Conference_Location

Curitiba

Print_ISBN

978-1-4244-5097-8

Type

conf

DOI

10.1109/ISAP.2009.5352938

Filename

5352938