Long‐term streamflow forecasting using artificial neural network based on preprocessing technique

摘要: Artificial neural network (ANN) combined with signal decomposing methods is effective for long‐term streamflow time series forecasting. ANN a kind of machine learning method utilized widely series, and which performs well in forecasting nonstationary without the need physical analysis complex dynamic hydrological processes. Most studies take multiple factors determining as inputs such rainfall. In this study, model depending only on historical data proposed. Various preprocessing techniques, including empirical mode decomposition (EMD), ensemble (EEMD) discrete wavelet transform (DWT), are first used to decompose into simple components different timescale characteristics, relation between these original at next step analyzed by ANN. Hybrid models EMD‐ANN, EEMD‐ANN DWT‐ANN developed study daily forecasting, performance measures root mean square error (RMSE), absolute percentage (MAPE) Nash–Sutcliffe efficiency (NSE) indicate that proposed better than EMD‐ANN models, especially high flow