Predviđanje proizvodnje termoelektrane pomoću neuralnih mreža

Abstract

With the power systems progress, its complexity grows and there is a greater need for good production organization and the need for information that would help in power system planning. As thermal power plants are responsible for covering the base load, the prediction of its production can be used significantly for good power system organization. This paper represents the use of artificial neural networks for the prediction of thermal plants power production. The data used for neural networks training are measurements of thermal power plant production in power systems of Serbia. Trained neural network can be used for prediction of production on yearly basis or for shorter, hourly or daily prediction.