Please use this identifier to cite or link to this item:
http://117.252.14.250:8080/jspui/handle/123456789/3046
Title: | 46-Estimation of suspended sediment concentration using Artificial Neural Networks |
Authors: | Kumar, A.R. Senthil Agarwal, Avinash Singh, R. D. Nema, R. K. |
Keywords: | Artificial Neural Networks |
Issue Date: | 2008 |
Publisher: | National Institute of Hydrology |
Abstract: | The assessment of the sediment volume transported by river water is very important in the design and management of water resources project. Several methods have been proposed to predict suspended sediment concentration based on the properties of flow and sediment. The equations proposed by the investigators for the estimation of sediment concentration have the limitations due to the simplification of important parameters and boundary conditions. Recently, neural networks approach has been applied in many areas of water resources due to its capability in representing any nonlinear processes by given sufficient complexity of the trained networks. In this paper, development of an Artificial Neural Network (ANN) model for predicting the suspended sediment concentration at the upstream of Bhakra reservoir is presented. The results of the ANN model for calibration indicated that the all range of sediment concentration values were simulated fairly well. Whereas the high range values of sediment concentration were slightly deviated from the observed values during the validation of the model. The performance of ANN model was compared with multiple linear regression model (MLR) and was found better than the MLR model. |
URI: | http://117.252.14.250:8080/jspui/handle/123456789/3046 |
Appears in Collections: | Proceedings of the National Seminar on Conservation and Restoration of lakes (CAROL-08), 16-17 October 2008 at Nagpur, Volume - II |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
46-Estimation of suspended sediment concentration using Artificial Neural Networks.pdf | 2.59 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.