Iran-Water Resources Research

Iran-Water Resources Research

Discharge Prediction by Comparing Artificial Neural Network With Fuzzy Inference System (Case study: Zayandeh rud River)

Document Type : Original Article

Authors
N Poustizadeh 1
N Najafi 2
1 Ph.D. Candidate, Water Structures., Chamran University, Ahvaz, Iran
2 M.Sc. Water Structures., Mazandaran University, Iran
Abstract
The Fuzzy Sets theory and the Artificial Neural Network are among the latest methods in Water Resources Engineering for systems dealing with complexity or ambiguity and lack enough data. The main advantage of these techniques over traditional methods is that they can investigate the effects of the available parameters on the process in a short time and with no need to cause complex mathematical equations. In this study time series of monthly discharge, temperature, and rainfall are used in the Artificial Neutral Network and  Fuzzy Inference System context in order to forecast the flow discharge for the upcoming months. Results of this research showed that the FIS gives better estimation than the ANN.
Keywords
Forecasting
River Flow
Fuzzy inference system
Artificial Neural Network
جمالی، س.، ابریشم­چی، الف و تجریشی، م. (1385). "ساخت مدل‌های پیش­بینی جریان رودخانه و بهره­برداری از مخزن با استفاده از سیستم استنباط فازی." مقاله دومین کنفرانس مدیریت منابع آب ایران.
کوره‌پزان دزفولی، الف. (1384). "تئوری مجموعه‌های فازی و کاربردهای آن در مدل‌سازی مسایل مهندسی آب." چاپ اول. انتشارات دانشگاه صنعتی امیرکبیر، 261ص.
 وانگ، ل. (1978)."سیستم­های فازی و کنترل فازی." ترجمه محمد تشنه لب و همکاران. انتشارات دانشگاه خواجه نصیر طوسی.
Hsu, k.-L., Gupta, H.V. and Sorooshnian, S. (1995), "Artificial neural networks modeling of rainfall-runoff process", Water Resources Research, 31(10), pp. 2517-2530.            
Mamdani, E.H. (1976), "Advances in linguistic synthesizes of fuzzy controllers", Journal of  Man mach, Stud., vol. 8, pp. 669-678.                                                           
 Ross, T.J. (1995), Fuzzy Logic with Engineering Applications, McGraw-Hill, Inc., USA.                                                  
 Salas, J.D., Markus, M. and  Tokar, A.S. (2000), "Streamflow forecasting based on artificial neural networks.In: Artifcial neural networks in hydrology", ed. Govindaraju, R.S. and Ramachandra, A., PP. 23-51. West Lafayette Indiana,U.S.A.
Sarangi, A. (2005), "Comparison of Artificial Neural Network and regression models for sediment loss prediction from Banha watershed in India", Journal ofAgricultural water management,  Pub. By Elsevier.    
Iran-Water Resources Research
Volume 7, Issue 2
Summer 2011
Pages 92-97

  • Receive Date 09 September 2009
  • Accept Date 08 December 2010