| Summary: | With electricity markets birth, electricity price volatility becomes one of the major concerns for their participants and in particular, for the producers. Whether or not to hedge, what type of portfolio is ade-quate, and how to manage that portfolio are important considerations for electricity market agents. To achieve that, load and electricity price forecast have a high impor-tance. This paper provides an approach applied to price range forecast. Making use of artificial neural networks (ANN), the methodology presented here has as main con-cern finding the maximum and the minimum System Mar-ginal Price (SMP) for a specific programming period, with a certain confidence level. To train the neural networks, probabilistic information from past years is used. To in-crease accuracy and turning ANN training more efficient, a K-Means clustering method is previously applied. Re-sults from real data are presented and discussed in detail.
|
|---|