Graphical Modeling for Selecting Input Variables of Short-term Load Forecasting

Abstract

This paper proposes a Graphical Modeling method for selecting input variables of short-term load forecasting in power systems. Short-term load forecasting plays a key role to smooth operation and planning such as economic load dispatching, unit commitment, etc. In addition, the deregulated power market players require more accurate prediction models for short-term load forecasting to maximize a profit and minimize the risk As a result, it is of importance to focus on the relationship between input and output variables. In this paper, a graphical modeling method is used to determine the appropriate input variables of ANN (artificial neural network) model in short-term load forecasting. It has advantage that more effective input variables are selected because of excluding the pseudo-correlation that gives more errors to the predicted value. The proposed method is tested for real data of short-term load forecasting.