A methodology for Short-term Electric Power Load Forecasting

Abstract

Energy consumption has been increasing steadily due to globalization and industrialization. As a result electricity load forecasting has gained vital importance in order to conserve energy and other resources. But due to the uncertain characteristics of forecasting methods, it is still one among the most difficult task to get implemented with accurate results. To predict the load, Bayesian Neural Network model based on the historical load and meteorological data of the given geographical region is presented in this article. To validate the performance of the model, meteorological and load consumption data in Kerala region over the period 2011–2012 have been used. Better accuracy and relatively shorter computing time assert that the proposed method can be used as an effective method for short-term load forecasting.