High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

Prediction of Runoff using Matlab

Author(s):

Prof. Shrivastav Mahesh B. A. Kayesth , Sir Bhavsinhji Polytechnic Institute, Bhavnagar, Gujarat

Keywords:

Rainfall-runoff Model, ANN, algorithm, simulation, prediction, time-series

Abstract

The present study aims to utilize an Artificial Neural Network (ANN) for modeling the rainfall runoff relationship of Machhundri catchment area located in Junagadh district, Gujarat, India. An Artificial Neural Network (ANN) methodology was employed to predict monthly runoff as a function of precipitation, temperature, evaporation losses, infiltration losses and humidity. The paper illustrates the applications of the feed forward network for the Runoff prediction with various algorithms and accordingly, different structures of ANNs were used and their efficiencies in terms of the mean squared error ‘MSE’, training and validation determination coefficients ‘R’ to select better predicted Runoff data were examined. The monthly hydrometric and climatic data of Machhundri Watershed in ANN were ranged from 1991 to 2016 and analyzed in order to calibrate the given models. Efficiencies of the Back-Propagation (BP), conjugate gradient (CG) and Levenberg-Marquardt (L-M) training algorithms are compared to improving the computed performances and 72 models were prepared to select a best model having mean square error ‘MSE’ nearer to zero and co-relation factor ‘R’ nearer to unity. The results revealed that the best model is composed of the feed-forward networks, trained by the Levenberg-Marquardt algorithm and considering only one hidden layer. The results extracted from the comparative study indicated that the Artificial Neural Network method is more appropriate and efficient to predict the river runoff than classical regression model. The ANN model provides a more systematic approach, reduces the length of calibration data, and shortens the time spent in calibration of the models.

Other Details

Paper ID: IJSRDV5I41523
Published in: Volume : 5, Issue : 4
Publication Date: 01/07/2017
Page(s): 1727-1730

Article Preview

Download Article