Previsão de carga no curtíssimo prazo utilizando redes neurais artificiais

Abstract

The problem of load forecasting is a very important component of strategic planning and operation of systems for generation, transmission and distribution of electric energy. Due to their degree of importance and appeal of this issue, this study highlights the relevant features, diferente techniques and methods which involve the problem of load forecasting in the very short term through studies in the literature. Additionally, this work focuses mainly on methods based on Artificial Neural Networks, from artificial intelligence precisely for its high degree of generalization from which he received wide use in scientific work in recent decades. Based on these methods, this paper will demonstrate, through a computer system, the results of the load forecasts and analysis for very short term, given the different parameters discussed.