Comparison of Neural Network and Transfer Function Methods for Estimating Suspended Sediment Load (Case Study: south of Hamedan Province)

طهماسبی پور, ناصر and رئیسي, مریم (2018) Comparison of Neural Network and Transfer Function Methods for Estimating Suspended Sediment Load (Case Study: south of Hamedan Province). In: The 13th National Conference on Watershed Management Science & Engineering of Iran and The 3rd National Conference on Conservation of Natural Resources and Environment, 3-4 October 2018, Ardabil, Iran.

[img] Text (مقایسه روش¬های شبکه عصبی و تابع انتقال در برآورد رسوب معلق)
wms97_UMA2018P157T12_FullPaperSubmission.pdf

Download (330kB)

Abstract

Estimation and prediction of sediment discharge is utilized in a wide range of issues such as engineering of rivers, design of dams, transmission of contamination, environmental effects, watershed management effects and many discussions of water resources. Therefore, in order to importance of estimation of the sediment load in many of studies of water resources, In this paper the transfer function and the neural network methods has been utilized because of estimation of the suspended sediment load in three hydrometric stations: Firooz Abad, Aran and Doab in south of Hamedan Province, and the accuracy of their forecasting has been compared with each other. The results of the comparison of these two above- mentioned methods shows that the transfer function method with SSE 2.73, 5.87 and 5.1 respectively for stations: Firooz Abad, Aran and Doab in comparison to the method of neural network with SSE 4.45, 6.48 and 5.32 has been more efficient. According to these results, the transfer function method which has more efficient in estimation of sediment in different conditions, in different basins is more recommended. Keywords: Neural Network, Transfer Function, Suspended Sediment and Hamedan Province

Item Type: Conference or Workshop Item (Poster)
Persian Title: مقایسه روش¬های شبکه عصبی و تابع انتقال در برآورد رسوب معلق
Persian Abstract: تخمین و پیش¬بینی دبی رسوب در طیف گسترده¬اي از مسائلی نظیر مهندسی رودخانه ها، طراحی سدها، انتقال آلودگی، تاثیرات زیست محیطی، تاثیرات آبخیزداري و بسیاري از بحث¬هاي منابع آب کاربرد دارد. لذا، در این مقاله از روش¬های تابع انتقال و مدل شبکه عصبی جهت تخمین مقدار رسوب معلق در سه ایستگاه هیدرومتری منتخب از جنوب استان همدان استفاده شده است و دقت پیش¬بینی آنها با یکدیگر مقایسه گردید. نتایج حاصل از مقایسه دو روش فوق نشان داد که مدل تابع انتقال با مجموع مربعات خطای 73/2، 87/5 و 1/ 5 به ترتیب در ایستگاه¬های فیروزآباد، آران و دوآب نسبت به روش شبکه عصبی با مجموع مربعات خطای 45/4، 48/6 و 32/ 5 کارآیی بیشتری داشته است. بر اساس این نتایج، روش سری زمانی تابع انتقال که تحت شرایط مختلف عملکرد بهتری دارد، جهت برآورد رسوب در حوضه¬های آبریز مختلف توصیه می¬گردد. کلمات کلیدی: شبکه عصبی، تابع انتقال، رسوب معلق، استان همدان
Subjects: Divisions > Conferences > The 13th National Conference on Watershed Management Science & Engineering of Iran and The 3rd National Conference on Conservation of Natural Resources and Environment
Conferences > The 13th National Conference on Watershed Management Science & Engineering of Iran and The 3rd National Conference on Conservation of Natural Resources and Environment
Divisions: Conferences > The 13th National Conference on Watershed Management Science & Engineering of Iran and The 3rd National Conference on Conservation of Natural Resources and Environment
Subjects > Conferences > The 13th National Conference on Watershed Management Science & Engineering of Iran and The 3rd National Conference on Conservation of Natural Resources and Environment
Date Deposited: 18 Oct 2018 12:48
Last Modified: 18 Oct 2018 12:48
URI: http://repository.uma.ac.ir/id/eprint/1024

Actions (login required)

View Item View Item