Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Modelling Total Suspended Solids in Vegetative Filter Strips using Artificial Neural Networks

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan

Citation:  Paper number  032079,  2003 ASAE Annual Meeting . (doi: 10.13031/2013.13770) @2003
Authors:   S. Andriyas, S. C. Negi, R. P. Rudra, S. X. Yang
Keywords:   Back propagation algorithm, Fuzzy c-means clustering algorithm, Multilayer perceptron, Radial Basis Function Network, Vegetative Filter Strips

Sediments and nutrients such as nitrogen and phosphorus, from cropland are carried by runoff to streams and lakes. Vegetative filter strips, being one of the best management practices, offer a feasible and economic solution to control the quality of downstream water. Artificial neural network models, as compared to hydrologic models, are liberal in terms of input patterns being given and aptly handle complicated modelling problems with ease, due to their capability of generalising the relationship between input and output variables. Therefore in this study, a radial basis function neural network using fuzzy c-means clustering algorithm and a multilayer perceptron using back propagation training algorithm have been evaluated, to predict the performance of the vegetative filter strips in terms of trapping efficiency of sediments. The multilayer perceptron using back propagation algorithm with 15 (R2=0.99) hidden units, with tanh and logsig, hidden and output activation functions, respectively, was found to work well for the application. The network was tested with 5% of test data out of 1316 total patterns.

(Download PDF)    (Export to EndNotes)