سال انتشار: ۱۳۸۷

محل انتشار: اولین کنفرانس نیروگاههای برق

تعداد صفحات: ۱۷

نویسنده(ها):

K. Afshar – EE Department, Imam Khomeini International University, Qazvin, Iran
N. Bigdeli –

چکیده:

Market data analysis and short term price forecasting in Iran electricity market as a market with payas-bid payment mechanism has been considered in this paper. The data analysis procedure includes bothcorrelation and predictability analysis of the most important load and price indices. The employed data are theexperimental time series from Iran electricity market in its real size and is long enough to make it possible totake properties such as non-stationarity of market into account. For predictability analysis, the bifurcationdiagrams and recurrence plots of the data have been investigated. The results of these analyses indicate existenceof deterministic chaos in addition to non-stationarity property of the system which implies short-termpredictability. In the next step, two artificial neural networks have been developed for forecasting the two price indices in Iran’s electricity market. The models’ input sets are selected regarding four aspects: the correlation properties of the available data, the critiques of Iran’s electricity market, a proper convergence rate in case ofsudden variations in the market price behavior, and the omission of cumulative forecasting errors. Thesimulation results based on experimental data from Iran electricity market are representative of good performance of the developed neural networks in coping with and forecasting of the market behavior, even in the case of severe volatility in the market price indices