期刊名称：Journal of Urban and Environmental Engineering
印刷版ISSN：1982-3932
电子版ISSN：1982-3932
出版年度：2009
卷号：3
期号：1
页码：1-6
语种：English
出版社：Universidade Federal da Paraíba
其他摘要：Without a doubt the carried sediment load by a river is the most important factor in creating and formation of the related Delta in the river mouth. Therefore, accurate forecasting of the river sediment load can play a significant role for study on the river Delta. However considering the complexity and non-linearity of the phenomenon, the classic experimental or physical-based approaches usually could not handle the problem so well. In this paper, Artificial Neural Network (ANN) as a non-linear black box interpolator tool is used for modeling suspended sediment load which discharges to the Talkherood river mouth, located in northern west Iran. For this purpose, observed time series of water discharge at current and previous time steps are used as the model input neurons and the model output neuron will be the forecasted sediment load at the current time step. In this way, various schemes of the ANN approach are examined in order to achieve the best network as well as the best architecture of the model. The obtained results are also compared with the results of two other classic methods (i.e., linear regression and rating curve methods) in order to approve the efficiency and ability of the proposed method.
关键词：River mouth; River Delta; Sediment load; Black box modeling; Artificial neural networks
其他关键词：River mouth; River Delta; Sediment load; Black box modeling; Artificial neural networks.
