This Electrical Engineering Undergraduate Project consists in the time series forecasting of the load of a special consumer, by Artificial Intelligence techniques. Several methods are available for predicting of the consumption of electrical energy. However, recurrent neural networks, especially Long Short-Term Memory networks, have been standing out in time series forecasting due to their long term memory capacity. The objective of this work is the development of a computational program that applies Long Short-Term Memory networks to predict demand, with the use of distance metrics for result evaluation.