Title : ( Application of the combination of hydrodynamic and artificial neural network approaches for river peak flow prediction )
Authors: Mohammad Taghi Dastorani , Nigel G. Wright ,
Full TextAbstract
The application of artificial neural networks for the correction of the outputs of a 1D hydrodynamic flow model in a semi arid catchment has been investigated in this study. A hydrodynamic model was constructed to predict flow at the outlet using time series data from upstream gauging sites as boundary conditions the results was not close enough to the actual values. Then the model was replaced by an ANN model but the results were not desirable. Finally the error of the model was predicted using a three-layer feedforward neural network model to optimise the outputs. This gave a significant improvement in the results. Due to suspension of flow gauging in one of the upstream sites, there is no data for this site for the last decade. To evaluate the adaption of models with this problem, all simulations were repeated but without data from the suspended site. A combination of these two techniques produced outputs that were more accurate than the results of the models individually
Keywords
, Artificial neural networks, Hydrodynamic models, River, Peak flow
@inproceedings{paperid:1059563,
author = {Dastorani, Mohammad Taghi and Nigel G. Wright},
title = {Application of the combination of hydrodynamic and artificial neural network approaches for river peak flow prediction},
booktitle = {Sixth international river engineering conference},
year = {2003},
location = {Ahvaz, IRAN},
keywords = {Artificial neural networks; Hydrodynamic models; River; Peak flow},
}
%0 Conference Proceedings
%T Application of the combination of hydrodynamic and artificial neural network approaches for river peak flow prediction
%A Dastorani, Mohammad Taghi
%A Nigel G. Wright
%J Sixth international river engineering conference
%D 2003
