One day-ahead price forecasting for electricity market of Iran using combined time series and neural network model

Price forecasts provide crucial information for generators. They plan bidding strategies for maximizing their own profits in the competitive electricity markets. Hence, generation companies (GENCOs) need precise price forecasting tools. This paper provides one highly accurate yet efficient tool for short term price forecasting based on combination of time series and artificial neural network methods (ANNs). First, input variables needed for neural network are determined by time series. This model relates the current price to the values of past prices. Second, neural network is used for one day a head price forecasting. Designed ANN based on feed-forward back propagation was trained and tested using year 2005 data from the electricity market of Iran. The results are tested with the extensive data sets, and good agreement is found between actual data and NN results. Results show that the proposed model forecasts prices with high accuracy for short term periods.