Simulation and Forecasting of River Flow Using Neural Network and a Fourier Series Model

Document Type : Original Article

Authors

1 Assistant Professor, Shahid Abbaspoor University, Iran

2 Water Resources Management Company, Iran

3 Jamab Consulting Engineers, Tehran, Iran

Abstract

This paper shows concurrent application of The Artificial Neural Network (ANN) model with Fourier Series ARIMA Model (FSAM).The FSAM model represent spectral analysis of precipitation of Barandooz river basin in west of Lake Urmia for simulation of river flow. Neural network are applied for simulation and forecasting of Barandooz river based on FSAM forecasts in the absence of precipitation data. Because temperature is one of the independent variables in the simulation model, generation of temperature data was carried out by twelve MLP Neural networks, one for each month. The analysis and results are presented in this paper. 
 

Keywords

References

افشار، ن. ر، فهمی، ه.(1376).گزارش پیش‌بینی بارش کل کشور. سازمان تحقیقات منابع آب.
 منهاج، م. ب. (1379). مبانی شبکه‌های عصبی. دانشگاه صنعتی امیر کبیر، جلد اول.
نیرومند، ح. ع. (1376). تحلیل سری‌های زمانی، دانشگاه فردوسی مشهد.
Howard, D. and Mark, B. (2001). User’s Guide for Neural Network Toolbox with Matlab. Version 6.1.
Kuo, L. H., Hoshin, V. G. and Sorooshian, S. (1993). Artificial Neural Network Modeling of the Rainfall-Runoff Process. Water Resources Research. 29(4): pp. 1185-1194.
Nguyen, T. D., Huynh, H. P. and Ashim, D. S. (1999). Neural Network Models For Flow Forecasting, Water SA 25(1): pp. 33-39.
Iran-Water Resources Research
Volume 2, Issue 1
June 2006
Pages 36-44

Files

History

  • Receive Date: 22 July 2003
  • Accept Date: 28 March 2005

Share

How to cite

Statistics