Search

This paper presents an approach for short-term electricity price and load forecasting using the artificial neural network (ANN) computing technique. The described approach uses the three-layered ANN paradigm with back-propagation. The publicly available data, acquired from the deregulated Victorian power system, was used for training and testing the ANN. The ANN approach based on similarity technique has been proposed according to which the load and price curves are forecasted by using the information of the days being similar to that of the forecast day. A Euclidean norm with weighted factors is used for the selection of similar days. Two different ANN models, one for load forecasting and another for price forecasting, have been proposed. Test results show that average price and load MAPEs for the year 2003 by using the ANN approach are obtained as 14.29% and 0.95%, respectively. MAPE values obtained from the price and load forecasting results confirm considerable value of the ANN based approach in forecasting short-term electricity prices and loads. [ABSTRACT FROM AUTHOR]

Abstract: In daily power markets, forecasting electricity prices and loads are the most essential task and the basis for any decision making. An approach to predict the market behaviors is to use the historical prices, loads and other required information to forecast the future prices and loads. This paper introduces an approach for several hour ahead (1–6h) electricity price and load forecasting using an artificial intelligence method, such as a neural network model, which uses publicly available data from the NEMMCO web site to forecast electricity prices and loads for the Victorian electricity market. An approach of selection of similar days is proposed according to which the load and price curves are forecasted by using the information of the days being similar to that of the forecast day. A Euclidean norm with weighted factors is used for the selection of the similar days. Two different ANN models, one for one to six hour ahead load forecasting and another for one to six hour ahead price forecasting have been proposed. The MAPE (mean absolute percentage error) results show a clear increasing trend with the increase in hour ahead load and price forecasting. The sample average of MAPEs for one hour ahead price forecasts is 9.75%. This figure increases to only 20.03% for six hour ahead predictions. Similarly, the one to six hour ahead load forecast errors (MAPE) range from 0.56% to 1.30% only. MAPE results show that several hour ahead electricity prices and loads in the deregulated Victorian market can be forecasted with reasonable accuracy. [Copyright &y& Elsevier]