Título: An Application of Elman Networks In Treatment and Prediction of Hydrologic Time Series
Autores: Tampelini, Leonardo G.; Boscarioli, Clodis; Peres, Sarajane M.; Sampaio, Silvio C.
Resumo: Brazil has an available hydrograph to build great dams. This shows that sophisticated runoff control systems with hydrological data prediction functionalities are necessary to deal with physical processes of high complexity and variability. A modeling of hydrological series according to conceptual methods is an expensive process and requires a lot of intervention from the experts. The application of Artificial Neural Networks is an alternative to capture the existing standards in hydrological time series, since it reduces such intervention and the cost of building a model. This paper presents the application of Elman Network for modeling of hydrological time series (imputation and prediction) through the construction of Rainfall- Runoff models and certifies its ability on generating reliable predictions of future values of river discharge based only on rainfall data.
Palavras-chave: Artificial Neural Network; Elman Network; Data Imputation; Time Series Prediction
Páginas: 9
Código DOI: 10.21528/lmln-vol9-no3-art1
Artigo em PDF: vol9-no3-art1.pdf
Arquivo BibTex: vol9-no3-art1.bib