Título: Previsao de Vazão com Séries Temporais Nebulosas para Bacias do Sistema Interligado Nacional
Autores: Tiago da Rocha Alves, Ivo Chaves da Silva Junior, Leonardo M. Honorio
Resumo: This study proposes a comparative analysis between the Soil Moisture Accounting Procedure (SMAP/ONS) and the fuzzy time series technique for water flow forecasting in Brazilian hydroelectric power generation systems. The reservoir crisis of 2021, caused by the worst drought since 1931, which affected the country’s energy sector and increased energy tariffs, served as a warning of the importance of water resource planning and management and how it is essential for electricity generation in Brazil. The SMAP/ONS model, currently used by the Brazilian National Electric System Operator (ONS), calculates water flow based on evapotranspiration and precipitation. In contrast, the fuzzy time series technique is a machine learning-based approach that uses fuzzy logic to handle uncertain data. Results show that the fuzzy time series technique presented competitive performance in most assessed cases.
Palavras-chave: Aprendizagem de máquina, Sistema Interligado Nacional, Previsao de Vazão, Séries Temporais Nebulosas
Páginas: 7
Código DOI: 10.21528/CBIC2023-066
Artigo em pdf: CBIC_2023_paper066.pdf
Arquivo BibTeX: CBIC_2023_066.bib