A Novel Approach to Forecast Day-Ahead Electricity Prices by Means of Neural Networks Using Groups of Similar Hours

作者：Menniti Daniele; Scordino Nadia^*; Sorrentino Nicola

来源：International Review of Electrical Engineering, 2012, 7(4): 5119-5133.

• Back Propagation
• Electricity Market Prices
• Forecasting
• Neural Network
• Time Series

Forecasting short-term electricity market prices has been the focus of several studies in recent years. Although various approaches have been examined, achieving sufficiently low forecasting errors has not always been possible. However, certain applications, such as demand-side management, do not require exact values for future prices, but utilize average values as the basis for making short-term scheduling decisions. With the aim of enhancing the accuracy of the day-ahead electricity price forecasting, field records of the Italian electricity market have been correlated so as to identify a number of blocks of hours, characterized by similarity of price. This paper thus proposes an approach to forecast the day-ahead electricity prices by means of a number of Neural Networks (NNs), with such a number being equal to the number of groups of hours with similarity of price, and with each NN forecasting the mean price over the hours belonging to the related group, as previously determined. Simulation results show that the fundamental and novel contribution of identifying, firstly, the membership of an hour to a particular group of hours, clustered according to their similarity of price, and then forecasting day-ahead prices for the resulting number of groups of hours performs Ably well if compared with other forecasting techniques, either on an hourly or on an average basis, in terms of mean absolute and absolute percentage errors, as well as of variance.

• 出版日期2012-8