Price Forecasting Model for Turkish Day-Ahead Electricity Market Using Neural Network

Day-ahead electricity market (DAM) price forecasts are crucial parameters for market participants to create their next – day generation plan. Accurate price forecasts help participant companies to increase their profit by shifting generation to high price occurred hours. In this study we developed a forecast model for Turkish energy market because country’s energy market mechanism hasn’t got price forecast module and there is no available accurate price forecasts. Architecture of model is based on feed forward back propagation neural network approach. Four year period real market data are used in train and test phases. Results of this study show that forecasts of proposed model have acceptable accuracy and the performance of the model strongly depends on the market demand and generation capacities.