Search

This paper presents a methodology for modeling typical power demand curves and shows the influence of climatic data as a deterministic factor for this characterization by using Artificial Neural Networks. The methodology presented was applied on real data, what made possible to obtain typical curves with a high degree of accuracy in forecasting demand curves. By means of the investigation carried out it can be seen the closest relationship between climatic data and electric power demand. Furthermore, a comparison between the proposed methodology and the kVAs method is presented. [ABSTRACT FROM PUBLISHER]