Streamflow simulation with an integrated approach of wavelet analysis and artificial neural networks

A loose type of wavelet neural network (WNN) model which utilizes the merits of the wavelet analysis method and artificial neural network is presented in this paper. The WNN model was applied to simulate the daily streamflow in the upper area of Nangao Reservoir at Shanwei City. The simulated streamflows with the WNN model were also compared to these simulated with back-propagation (BP) neural networks model for evaluating the performance of the WNN model. The numerical experiment shows that the simulation results with the WNN model are more accurate than these simulated with the BP model. The results also indicate that this method is feasible and effective for hydrological forecasting.