Application of Artificial Neural Networks in the Prediction of Photovoltaic Power Generation in Northeastern Brazil

• Autor(a) principal: Cunha, Henrique Queiroz
• Data de Publicação: 2021
• Outros Autores: Sobel, Leonardo Farias
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Revista de Engenharia e Pesquisa Aplicada
• Texto Completo: http://revistas.poli.br/index.php/repa/article/view/1767

The generation of energy through sustainable sources is becoming increasingly important in the current global market configuration. In Brazil, the Northeast region, due to its geographical location and climatic characteristics, has a best stability in the production of solar energy throughout the year. For this reason, it stands out among the other regions and has the largest installed capacity of solar plants in the country. Thus, this paper aims to evaluate the use of artificiais neurais networks as a method for predicting the generation of photovoltaic energy in the Brazilian Northeast. For this, the climatic and energy generation data available in the city of Bom Jesus da Lapa/ Bahia were collected and pre-processed. And, after processing, the prediction results obtained using the Multilayer Perceptron (MLP) and Long Short-Term Memory (LSTM) networks were compared with those of the classical linear ARIMA (Autoregressive Integrated Moving Average) method and showed good performance. Finally, the good result obtained by neural networks as tools for photovoltaic solar energy prediction is observaded.