Data mining of electricity price forecasting with regression tree and normalized radial basis function network

This paper proposes a new method for electricity price forecasting. The proposed method is based on the regression tree and NRBFN (Normalized Radial Basis Function Network) of ANN. The former is used to evaluate if-then rules and classify input data into some clusters. The latter is employed to calculate more accurate predicted values. The regression tree is one of data-mining techniques that extract if-then rules from database. NRBFN is an extension of RBFN (Radial Basis Function Network) that improves the generalization ability of RBFN. The effectiveness of the proposed method is demonstrated for real data of on-step ahead electricity price forecasting.

I. Introduction

IN recent years, the deregulation of power systems is widely spread in the world. The power market becomes more competitive so that the market players are interested in the maximization of profit and the minimization of risk. As a result, electricity price forecasting is one of the most important tasks under new environment of the deregulated power systems. It is related to the complicated nonlinearity due to the load demand, weather conditions such as temperature, amount of available generations, etc. The volatility of electricity price is much higher than the foreign exchange, the interest rate, the stock prices, etc. Also, it has a feature that the electricity is not stored unlike other commodities. As a result, the behavior is hard to predict. The conventional studies on electricity price forecasting may be classified into the following: (i)

statistical model [1], [2], [6]

(ii)

artificial neural networks (ANN) [3]–[8]