Subscribe Now Subscribe Today
Science Alert
FOLLOW US:     Facebook     Twitter
Curve Top
Journal of Environmental Science and Technology
  Year: 2016 | Volume: 9 | Issue: 3 | Page No.: 257-267
DOI: 10.3923/jest.2016.257.267
Calibrating Conceptual Rainfall Runoff Models using Artificial Intelligence
Mrad Dounia, Djebbar Yassine and Hammar Yahia

Rainfall runoff models are highly useful for water resources planning and development. In the present study, an effort has been made to develop three types of artificial intelligence techniques (genetic algorithms, fuzzy logic and artificial neural network) based rainfall runoff GR2M prediction model using current monthly rainfall, potential evapotranspiration and river basin area and give an output monthly runoff. The aim of this study is to evaluate the objective function between these three intelligence techniques in the Medjerda river basin, north east of Algeria. To do so, the mathematical model of GR2M is improved in MATLAB/Simulink and the proposed intelligence techniques are used. First, the offline GA setting is used to optimize the GR2M parameters. Second, technical intelligence, FL and ANN tuning online are used to consign to regulate by an adaptative control of the GR2M parameters. The GR2M model presented in our study with these proposed artificial intelligence techniques have been simulated in MATLAB/Simulink®. The performance of the model was evaluated qualitatively and quantitatively by visual observation and employing various statistical indices viz., correlation coefficient, root mean square error, coefficient of efficiency and volumetric error. The results showed that the neural network (ANN) is an effective algorithm to forecast rainfall runoff relation more accurately than the other techniques.
 [Fulltext PDF]   [Fulltext HTML]   [XML: Abstract + References]   [References]   [View Citation]  [Report Citation]
How to cite this article:

Mrad Dounia, Djebbar Yassine and Hammar Yahia, 2016. Calibrating Conceptual Rainfall Runoff Models using Artificial Intelligence. Journal of Environmental Science and Technology, 9: 257-267.

DOI: 10.3923/jest.2016.257.267






Curve Bottom