System identification to forecast electricity loads

Abstract

Forecasting electricity load is an important economic problem. For an electric load forecasting task with a prediction horizon of 1–31 days ahead we decide to develop a simple model based on a system identification using the ARX model. The linear structure of the ARX model is concerned with the number of time lags of the model and coefficients. The determination of the system order is very important in the performance of the estimated model which is estimated using the training data collected by the Eastern Slovakian Electricity Corporation during the period of 1997–1998. The model is validated by using real data from the first 20 days of December 1998. For final application, the model is retrained and the requested forecasts are then performed