Comparison of advanced learning algorithms for short-term load forecasting

Article type: Research Article

Authors: Kodogiannis, V.S.

Affiliations: Computer Science Department, University of Westminster, Harrow HA1 3TP, UK. Tel.: +44 20 79115000, ext. 4250; Fax: +44 20 79115906; E-mail: kodogiv@wmin.ac.uk

Abstract: Neural Networks are currently finding practical applications, ranging from 'soft' regulatory control in consumer products to accurate modelling of non-linear systems. This paper presents the development of improved neural networks based on short-term electric load forecasting models for the Power System of the Greek Island of Crete. Several approaches including radial basis function networks, dynamic neural networks and fuzzy-neural-type networks have been proposed and discussed in this paper. Their performances are evaluated through a simulation study, using metered data provided by the Greek Public Power Corporation. The results indicate that the load forecasting models developed provide more accurate forecasts compared to the conventional backpropagation network forecasting models. Finally, the embedding of the new models capability in a modular-constructed forecasting system is presented.

Journal: Journal of Intelligent and Fuzzy Systems, vol. 8, no. 4, pp. 243-259, 2000