Due to the energy turnaround in Germany and the associated Renewable Energy Sources Act, the share of renewable energy sources on the power market has increased continually in the recent years. As a consequence of this growth, electricity prices declined at the energy exchange forcing conventional energy generating companies to face new challenges. Hence, an accurate and target-oriented prediction of future price movements plays an important role in regard to risk management and planning of these companies. This thesis deals with the forecasting of electricity prices using artificial neural networks at the European Energy Exchange (EEX) based in Leipzig. In contrast to traditional statistical forecasting models, artificial neural networks are able to "learn from their experience". Even if the underlying data structure is nonlinear, too complex or not known, artificial neural networks are able to adapt the underlying pattern from the data fed into the network. Since electricity prices follow a nonlinear structure, an artificial neural network model is chosen to forecast future electricity prices. In particular, the artificial neural network model proposed by Pao is used in this thesis and applied to day ahead electricity prices of the spot market at the EEX. On the one hand, data from November 2002 to October 2005 and on the other hand data from 2010 to 2012 is considered to analyze the prediction accuracy and to test the functionality of the model in a changing market environment and regulatory conditions. Furthermore, comparison between daily and hourly day ahead forecasting is carried out. In order to evaluate the performance of the model, a cross validation method is used.
«
Due to the energy turnaround in Germany and the associated Renewable Energy Sources Act, the share of renewable energy sources on the power market has increased continually in the recent years. As a consequence of this growth, electricity prices declined at the energy exchange forcing conventional energy generating companies to face new challenges. Hence, an accurate and target-oriented prediction of future price movements plays an important role in regard to risk management and planning of thes...
»