Electricity consumption forecasting using fuzzy time series

Bolturk E., Öztayşi B., Sari İ.

13th IEEE International Symposium on Computational Intelligence and Informatics, CINTI 2012, Budapest, Hungary, 20 - 22 November 2012, pp.245-249 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/cinti.2012.6496768
  • City: Budapest
  • Country: Hungary
  • Page Numbers: pp.245-249
  • Keywords: Electricity demand forecasting, fuzzy time series, deseasonalization, ALGORITHM, DESIGN, MODEL
  • Istanbul Technical University Affiliated: Yes

Abstract

Electricity is an important issue in energy management. Because of characteristic of electricity's non-storability, management of electricity is very valuable process. In this study, a Turkish company's two years electricity consumption have examined in order to predict possible electricity consumption. We used fuzzy time series with Singh's Method. Further, with the same data we compared total values of three periods (Unified consumption amount) and total consumption forecasts with their actuals. Finally we decide that which way is more accurate with root-mean-square deviation evaluation. © 2012 IEEE.