GALINDO, L. F. M.; http://lattes.cnpq.br/7276983053325255; GALINDO, Luan Fábio Marinho.
Résumé:
This work aims to use Data Science techniques and Machine Learning algorithms to predict the electricity demand in Brazil from January to March 2023. Electricity demand prediction techniques are fundamental for the operational logistics of the National Electric System, in a way that the management of those resources are handled in a more efficiently way and reduces waste. In view of the above, this work has the specific objectives of collecting and processing a database for use in training, carrying out the training itself and then analyze the results. For this, the Python programming language will be used, together with the Application Programming Interface (API) Tensor Flow, in the Anaconda development platform with the Integrated Development Environment (IDE) Spyder. The database used will be obtained by the National Electric System Operator (ONS), by the Electric Energy Commercialization Chamber (CCEE) and by Yahoo! Finances. Finally, this work will demonstrate the potential of using Machine Learning techniques to forecast electricity demand, showing how the use of historical data contributes to a better management of the National Electric System.