Short term load forecasting using genetic algorithm and neural networks

Author

Heng, Edmund T H ; Srinivasan, Dipti ; Liew, A.C.

Author_Institution

Dept. of Electr. Eng., Nat. Univ. of Singapore, Singapore

Volume

2

fYear

1998

fDate

3-5 Mar 1998

Firstpage

576

Abstract

This paper presents an artificial neural network (ANN) model trained by a genetic algorithm (GA) for short term load forecasting. Genetic algorithms (GAs) seek to solve optimization problems using the methods of evolution, specifically survival of the fittest. The software, Genehunter from Ward Systems Group was used to build an ANN model capable of forecasting one-day ahead hourly loads for weekdays and weekends. The proposed model is a three-layered feedforward backpropagation network. The results indicate that this model can achieve greater forecasting accuracy than the traditional statistical model. This ANN model has been implemented on real load data for the year 1995

Keywords

backpropagation; feedforward neural nets; genetic algorithms; load forecasting; multilayer perceptrons; power system analysis computing; Genehunter software; Ward Systems Group; artificial neural network; genetic algorithm; one-day ahead hourly loads; optimization; real load data; short term load forecasting; statistical model; survival of the fittest; three-layered feedforward backpropagation network; weekdays; weekends; Artificial neural networks; Backpropagation; Expert systems; Genetic algorithms; Input variables; Load forecasting; Neural networks; Predictive models; Transfer functions; Weather forecasting;

fLanguage

English

Publisher

ieee

Conference_Titel

Energy Management and Power Delivery, 1998. Proceedings of EMPD '98. 1998 International Conference on

Print_ISBN

0-7803-4495-2

Type

conf

DOI

10.1109/EMPD.1998.702749

Filename

702749