Title: Price volatility forecasting using artificial neural networks in emerging electricity markets
Authors: Ahmad F. Al-Ajlouni; Hatim Y. Yamin; Ali Eyadeh
Addresses: Hijjawi Faculty for Engineering Technology, Communication Engineering Department, Yarmouk University, Irbid 21163, Jordan. ' Electrical Engineering Department, College of Engineering, King Saud University, Majmaah-Riyadh, Saudi Arabia. ' Hijjawi Faculty for Engineering Technology, Communication Engineering Department, Yarmouk University, Irbid 21163, Jordan
Abstract: In the adaptive short-term electricity price forecasting, it may be premature to rely solely on the hourly price forecast. The volatility of electricity price should also be analysed to provide additional insight on price forecasting. This paper proposes a price volatility module to analyse electricity price spikes and study the probability distribution of electricity price. Two methods are used to study the probability distribution of electricity price: the analytical method and the ANN method. Furthermore, ANN method is used to study the impact of line limits, line outages, generator outages, load pattern and bidding strategy on short term price forecasting, in addition to sensitivity analysis to determine the extent to which these factors impact price forecasting. Data used in this study are spot electricity prices from California market in the period which includes the crisis months where extreme volatility was observed.
Keywords: probability distribution; competitive markets; ANNs; price forecasting; volatility analysis; price volatility; price forecasting; artificial neural networks; electricity markets; emerging markets; electricity prices; line limits; line outages; generator outages; load patterns; bidding strategy; short term pricing.
DOI: 10.1504/IJPEC.2012.048043
International Journal of Power and Energy Conversion, 2012 Vol.3 No.3/4, pp.157 - 183
Accepted: 08 Apr 2011
Published online: 07 Aug 2014 *