مدل سازی پیشی بینی میزان رسوب رودخانه قلعه رودخان با استفاده از شبکه عصبی LSTM
محورهای موضوعی : رسوب
محبوبه شادابی بجند
1
(دانشجوی دکتری گروه مهندسی آب، واحد لاهیجان، دانشگاه آزاد اسلامی، لاهیجان، ایران. *(مسوول مکاتبات))
ابراهیم امیری
2
(استاد گروه مهندسی آب، واحد لاهیجان، دانشگاه آزاد اسلامی، لاهیجان، ایران.)
کلید واژه: استان گیلان, رسوب لحظه ای, رودخانه قلعه رودخان, شبکههای عصبی بازگشتی, LSTM.,
چکیده مقاله :
زمینه و هدف: برآورد مناسب از میزان رسوب جاری شده در رودخانهها به عنوان مبنای داده¬ای برای بسیاری از طرحها و فرآیندهای مهندسی رودخانه دارای اهمیت است. رودخانه قلعه رودخان یکی از حوزه¬های آبی بسیار مهم در غرب استان گیلان می باشد. رودخانه قلعه رودخان از دو شاخه (حیدرآلات) و (نظر آلات) تشکیل شده است. از همین رو، هدف از انجام این پژوهش، مدل-سازی پیشی بینی میزان رسوب رودخانه قلعه رودخان با استفاده از شبکه عصبی حافظه طولانی کوتاه مدت (LSTM) است. روش بررسی: در این تحقیق از آمار دبی ـ رسوب ثبت شده مربوط به دوره آماری سال 1381 تا 1395 استفاده شده است. این آمار شامل دبی لحظه ای روزانه به مترمکعب بر ثانیه و رسوب روزانه لحظه ای به تن در روز است که همزمان اندازه گیری شده اند. متغیرهای تحت بررسی در مدلسازی پیش بینی مستلزم ايجاد يک شبکه عصبي مصنوعي، وجود يک سري داده، به منظور مدلسازي در اين شبکه مي باشد. یافته ها: دقت پیش بینی های انجام شده با سه معیار خطا بررسی شد. سه معیار مورد بررسی به ترتیب AFE، FFE و n-AFE هستند. بحث و نتیجهگیری: نتایج به دست آمده نشان داد که از میان معیارهای مورد بررسی معیار FFE همبستگی میان خروجی مدل و داده¬های اندازه¬گیری شده رسوب مناسب می باشد. در نتیجه مدل LSTM دارای دقت مناسب برای پیش بینی مقدار رسوب دو رودخانه قلعه رودخان می باشد.
Background and Objective: Proper estimation of the amount of sediment flowing in rivers is important as a data base for many river engineering designs and processes. Qaleh Rudkhan River is one of the most important water basins in the west of Gilan province. The most important branches of the basin are two branches named Gasht Rudkhan and Ghaleh Rudkhan. The river (Qaleh Rudkhan) is made up of two branches (Heydaralat) and (Nazaralat). Therefore, the purpose of this study was to model the prediction of sediment rate in Qaleh Rudkhan River using long short-term memory neural network (LSTM). Material and Methodology: In this research, the recorded Debi-sediment statistics related to the statistical period of 1381 to 1395 has been used. These statistics include daily instantaneous Debi in cubic meter per second and daily instantaneous sediment in ton per day, which are measured simultaneously. The data used to model the artificial neural network are Debi-sediment values the accuracy of the predictions was evaluated with three error criteria. Findings: The three criteria considered are AFE, FFE and n-AFE, respectively. Discussion and Conclusion: Among these criteria, the FFE criterion showed that the correlation between the model output and the measured sediment data is appropriate. As a result, the LSTM model has the appropriate accuracy to predict the amount of sediment in the two rivers of Qala-e-Rudkhan.
1. Oliveira N, Cortez P, Areal N. The impact of microblogging data for stock market prediction: Using Twitter to predict returns, volatility, tradingvolume and survey sentiment indices. Expert Systems with Applications, 2017; 73: 125-144.
2. Batny N, Golmaee S.H, Zia Tabar Ahmadi M.Kh. 2015. The study of sediment transport and the changes of river bed using Gstars3 Mathematical model (Case study: Gaveh Roud River). J. of Water and Soil Conservation, Vol. 22(1), (in Persian).
3. Hezbavi, Z., Sadeghi, S.H.R. And Younesi H.A. 2012. Analysis and evaluation of the effectiveness of runoff components from the application of different levels of polyacrylamide. Journal of Soil and Water Resources Protection, 2 (2): 1-14. (In Persian)
4. Behzadfar, M., Sadeghi, S.H.R., Khanjani, M.J. And Hezbavi, Z. 1391. Influence of runoff production and sediment of soils under freezing-thaw cycle under rain simulation conditions. Journal of Soil and Water Resources Protection, 2 (1): 13-23. (In Persian)
5. Karami, A., Homaee, M., Neyshabouri, M.R., Afzalinia, S. and Basirat, S. 2012. Large scale evaluation of single storm and short/long term erosivity index models. Turkish Journal of Agriculture and Forestry, 36: 207-216.
6. Dehghani, A.A., Zanganeh, M.E., Mosaed, A., and Kohestani N. 2009. Comparison of Suspended Sediment Estimation by Artificial Neural Network and Sediment Rating Curve Methods (Case Study: Doogh River in Golestan Province). J. Agric. Sci. Natur. Resour., Vol. 16 (Special issue 1-a), 2009. (In Persian)
7. Ebrahimi Mohammadi, SH., Sadeghi, S.H., and Chapi, K. 2012. Analysis of runoff, suspended sediment and nutrient yield from different tributaries to Zarivar lake in event and base flows. Artery of water and soil protection, second year, first issue, autumn 1. (In Persian)
8. Vali, A., Ramesht, M.H., Seif, A., and Ghazavi, R. 2011. Comparison of the efficiency of artificial neural network and regression models for prediction Sediment load during a case study of Samandgan watershed. Journal of Geography and Environmental Planning, Volume 22, Number 44, Issue 4, Winter 2011. (In Persian)
9. Eshghi, P., Farzadmehr, J., Dastoran, M.T., and Arabasadi, Z. 2016. The Effectiveness of Intelligent Models in Estimating the River Suspended Sediments (Case Study: Babaaman Basin, Northern Khorasan). Journal of Watershed Management Research Vol. 7, No. 14, Autumn and Winter 2016. (In Persian)
10. Kakaei Lafdani, E.,Moghaddam Nia, A. and Ahmadi, A. 2013. Daily suspended sediment load prediction using artificial neural networks and support vector machines. Hydrology, 478: 50-62. (In Persian)
11. Khanchoul, K., Altschul, R., and Assassi, F. 2010. Estimating suspended sediment yield , sedimentation controls and impacts in the Mellah Catchment of Northern Algeria. Arab. J. Geosci. 2: 3. 257-271.
12. Platt, J. 2000. Fast Training Support Vector Machine Using Sequential Minimal Optimization. http://www.research.microsoft.com/_jplatt. 41-65.
13. Yosefi, M., A. Talebi and R. Poorshariaty. 2014. Application of Artificial Intelligence in Water and Soil Sciences. Yazd University publication, Yazd, Iran, 516 pp. (In Persian)
14. Tabatabaei, M., K. Solaimani, M. Habibnejad Roshan and A. Kavian. 2014. Estimation of Daily Suspended Sediment Concentration Using Artificial Neural Networks and Data Clustering by Self Organizing Map (Case Study: Sierra Hydrometry Station- Karaj Dam Watershed). Journal of Watershed Management, 5: 98-116. (In Persian)
15. Falamaki, A., M. Eskandari, A. Baghlani and A. Ahmadi. 2013. Modeling Total Sediment Load in Rivers Using Artificial Neural Networks. Journal of Water and Soil Conservation, 2: 13-26. (In Persian)
16. Akbari, Z. and A. Talebi. 2010. Estimation of Suspended Sediment Using Regression Decision Trees Method (Case Study Ilam Dam Basin Science and Technology of Agriculture and Natural Resources Journal, 17: 109-121. (In Persian)
17. [17]. Dastorani, M., Kh. Azimi Fashi, A. Talebi and M. Ekhtesasi. 2012. Suspended Sediment Estimation Using Artificial Neural Network (Case Study: Jamyshan watershed in Kermanshah). Journal of Watershed Management, 3: 61-74. (In Persian)
18. Toloei, S., D. Hossenzadeh, M. Ghorbani, A. Fakhrefard and F. salmasi. 2011. Estimate Temporal and Spatial Suspended load river AJICHAI with Use from Geostatistics and Artificial neural Network. Issue Science Water and Soil, 21: 12-25. (In Persian)
19. Kisi, O. and Shiri, J. 2012. River Suspended Sediment Estimation by Climate Variables Implication: Comparative Study among Soft Computing Techniques. Computer and Geosciences, 43: 73-82.
20. Hussain D, Khan AA. Machine learning techniques for monthly river flow forecasting of Hunza River, Pakistan. Earth Science Informatics, 2020; DOI: 10.1007/s12145-020-00450-z.
21. Dou M, Qin C, Li G, Wang C. Research on Calculation Method of Free flow Discharge Based on Artificial Neural Network and Regression Analysis. Flow Measurement and Instrumentation, 2020; 72: 102-123.
22. Amiri, E., Naderi Dyzgahi, M.f., and Baygan,A. 2015. Investigation of discharge changes and water quality of Rudkhan Castle. National Association of Architects of Iran. International Conference on Civil Engineering, Architecture and Urban Planning in the Third Millennium. Tehran - July 2015. (In Persian)
23. Ahmadi P, Arefee H, Kardan N. 2020. Prediction of average monthly discharge of Karun river using GRU- LSTM combined method. Journal of Echo Hydrology, Volume 7. Issue 3. Fall 1399. Page 633-619.
24. Razavizadeh S, Kavian A.A, Vafakhah M. 2014, Estimation of suspended sediment discharge flow using the best network structure Artificial nerve in Taleghan watershed Journal of Agricultural Science and Technology and Natural Resources, Soil and Water Sciences / Year 18 / Issue Sixty-Eight / Summer 1 (In Persian)
