ارزیابی کارایی انواع تبدیل موجک در مدلسازی ترکیبی موجک- شبکه عصبی مصنوعی برای پیش‌بینی جریان ماهانه رودخانه (مطالعه موردی: رودخانه کارده)

نوع مقاله : مقالات پژوهشی

نویسندگان

• عاطفه کاظمی چولانک
• فرشته مدرسی
• ابوالفضل مساعدی

گروه علوم و مهندسی آب، دانشکده کشاورزی، دانشگاه فردوسی مشهد، مشهد، ایران

https://doi.org/10.22067/jsw.2024.86414.1371

چکیده

رواناب پدیده‎ای مهم در چرخه هیدرولوژیکی است، از این رو پیش‌بینی میزان رواناب رودخانه برای اهدافی نظیر برنامه‌ریزی فعالیت‌های کشاورزی، پیش‌بینی سیلاب و تأمین آب مصرفی حائز اهمیت است. پیچیده بودن مدل‏های فیزیکی یکی از دلایلی است که باعث شده پژوهشگران به مدل‏های داده‌مبناء و مبتنی بر هوش مصنوعی روی آورند. وجود تغییرات آماری در داده‌ها سبب می‌شود که مدل‌سازی جریان رودخانه با مدل‌های داده‌مبناء با مشکلاتی در فرآیند یادگیری مدل همراه باشد. لذا لازم است با مدل‌سازی تلفیقی، دقت پیش‌بینی جریان ارتقاء یابد. هدف تحقیق حاضر، ارزیابی کارایی انواع موجک‌های گسسته و پیوسته در مدل ترکیبی موجک-شبکه عصبی (WANN) برای پیش‌بینی جریان ماهانه رودخانه کارده در ایستگاه ورودی به سد کارده است. بدین منظور، دو موجک گسسته Haar و Fejer-Korovkin2 و دو موجک پیوسته Symlet3 و Daubechies2 در ترکیب با مدل ANN مورد ارزیابی قرار گرفت. بررسی داده‌های هواشناسی و هیدرومتری در یک دوره 30 ساله (1370-1399) نشان داد که جریان ماهانه در دو گام زمانی T-1 و T-2 بهترین متغیرهای پیش‌بینی‌کننده (در سطح اطمینان 95%) بودند. آنالیزهای ترکیبی در سه سطح تجزیه انجام و کارایی مدل‌ها با روش صحت‌سنجی متقاطع در4 سطح مورد ارزیابی قرار گرفت. نتایج نشان داد که مدل‌های ترکیبی دارای دقت بالاتری نسبت به مدل ANN بودند و مدل ترکیبی پیشنهادی Symlet3-ANN در سطح 3، نتایج بهتری نسبت به سایر مدل‎ها ارائه داد، بطوری‌که شاخص‌های R، RMSE و NSE در بخش واسنجی به‌ترتیب 90/0، 25/0 و 81/0 و در بخش صحت‌سنجی به‌ترتیب 85/0، 30/0 و 62/0 بود. همچنین ملاحظه شد دقت نتایج در سطح دو و سه تفاوت معناداری ندارند و بهتر است جهت کاهش مؤلفه‏های ورودی به مدل ANN و کاهش زمان اجرای مدل، تجزیه در سطح دو انجام شود.

کلیدواژه‌ها

• تبدیل موجک
• صحت سنجی متقاطع
• مدل هیبریدی
• موجک پیوسته
• موجک گسسته

موضوعات

• علوم آب

عنوان مقاله [English]

Assessment of the Performance of Various Wavelet Transforms in Combined Wavelet-neural Network Modeling for Monthly River Flow Prediction (Case Study: Kardeh Watershed)

نویسندگان [English]

• A. Kazemi Choolanak
• F. Modaresi
• A. Mosaedi

Water Science and Engineering Department, Faculty of Agriculture, Ferdowsi University of Mashhad, Mashhad, Iran

چکیده [English]

 
Introduction
Predicting river flow is one of the most crucial aspects in water resources management. Improving forecasting methods can lead to a reduction in damages caused by hydrological phenomena. Studies indicate that artificial neural network models provide better predictions for river flow compared to physical and conceptual models. However, since these models may not offer reliable performance in estimating unstable data, using preprocessing techniques is necessary to enhance the accuracy and performance of artificial neural networks in estimating hydrological time series with nonlinear relationships. One of these methods is wavelet transformation, which utilizes signal processing techniques.
 
Materials and Methods
In this study, to evaluate the efficiency of discrete and continuous wavelet types in the Wavelet-Artificial Neural Network (WANN) hybrid model for monthly flow prediction, a case study was conducted on the Kardeh Dam watershed in the northeast of Iran, serving as a water source for part of Mashhad city and irrigation downstream agricultural lands. Monthly streamflow estimates for the upstream sub-basin of the Kardeh Dam were obtained from the meteorological and hydrometric stations' monthly statistics over a 30-year period (1991-2020). The WANN model is a hybrid time series model where the output of the wavelet transform serves as a data preprocessing method entering an artificial neural network as the predictive model. The combination of wavelet analysis and artificial neural network implies using wavelet capabilities for feature extraction, followed by the neural network to learn patterns and predict data, potentially enhancing the models' performance by leveraging both methods. The 4-fold cross-validation method was employed for the artificial neural network model validation, where the model underwent validation and accuracy assessment four times, each time using 75% of the data for training and the remaining 25% for model validation. The final results were presented by averaging the validation and accuracy results obtained from each of the four model runs. To evaluate and compare the performance of the models used in this study, three evaluation indices, Nash-Sutcliffe Efficiency (NSE), Root Mean Square Error (RMSE), and Pearson correlation coefficient (R), were employed.
 
Results and Discussion
The analysis of meteorological and hydrometric data in this study revealed that monthly streamflow in two time steps, T-1 and T-2, were the most effective predictive variables. Each of the two runoff variables of the previous month (Qt-1) and the previous two months (Qt-2) were analyzed by each of the Haar and Fejer-Korovkin2 discrete wavelet transforms and the two continuous Symlet3 and Daubechies2 wavelets at three levels. The results of each level of decomposition was given as input to the ANN model. The presented results at each decomposition level indicated that hybrid models could accurately predict lower flows compared to the single ANN model, and the estimation of maximum values also significantly improved in the hybrid models. Among the wavelets used, Haar wavelets exhibited the weakest performance, and the less commonly employed Kf2 wavelet showed a moderate performance. Since the Haar and Fk2 wavelets, with their discrete structure, did not perform well in decomposing continuous monthly streamflow data, continuous wavelet models outperformed discrete wavelet models. The hybrid models, combining wavelet analysis and artificial neural networks, demonstrated up to an 11% improvement over the performance of the single neural network model.
 
Conclusion
Streamflow is a crucial element in the hydrological cycle, and predicting it is vital for purposes such as flood prediction and providing water for consumption. The objective of this research was to evaluate the performance of different types of discrete and continuous wavelet models at various decomposition levels in enhancing the efficiency of artificial neural network (ANN) models for streamflow prediction. Since climate and watershed characteristics can influence the nature of data fluctuations and, consequently, the results of the wavelet model decomposition, choosing an appropriate wavelet model is essential for obtaining the best results. Considering the existing variations in the results of different studies regarding the selection of the best wavelet type, it is suggested to use both continuous and discrete wavelet types in modeling to achieve the best predictions and select the optimal results. Given that a lower number of input variables in neural network models lead to higher accuracy in modeling results, it is recommended to perform decomposition at a two-level depth to reduce input components to the neural network model, thereby reducing the model execution time.

کلیدواژه‌ها [English]

• Artificial Neural Network
• Continuous wavelet
• Cross-validation
• Discrete wavelet
• Hybrid model
• Wavelet transform

©2024 The author(s). This is an open access article distributed under Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source.

مراجع

1. Abda, Z., Chettih, M., & Zerouali, B. (2021). Assessment of neuro-fuzzy approach based diferent wavelet families for daily fow rates forecasting. Journal of Model Earth Syst Environent, 7, 1523–1538. https://doi.org/ 1007/s40808-020-00855-1
2. Adamowski, J.F., & Sun, K. (2010). Development of a coupled wavelet transform and neural network method for flow forecasting of non-perennial rivers in semi-arid watersheds. Journal of Hydrology, 390(1), 85-91. https:// doi.org/1016/j.jhydrol.2010.06.033
3. Ahmadi, F., & Maddah, M.A. (2020). Development of wavelet-Kstar algorithm hybrid model for the monthly precipitation prediction (case study: synoptic station of shvaz). Journal of Soil and Water Research, 52(2), 409-420. (In Persian). https://doi.org/22059/IJSWR.2021.314110.668808
4. Ahmadi, M., Moeini, A., Ahmadi, H., Motamedvaziri, B., & Zehtabiyan, G.R. (2019). Comparison of the performance of SWAT, IHACRES and artifcial neural networks models in rainfall-runof simulation (case study: Kan watershed, Iran). Journal of Physics and Chemistry of the Earth, 111, 65–77. https://doi.org/ 1016/j.pce.2019. 05.002
5. Ahooghalandari, M., Khiadani, M., & Kothapalli, G. (2016). Assessment of Artifcial Neural Networks and IHACRES models for simulating streamflow in Marillana catchment in the Pilbara, Western Australia. Austr Journal Water Resource, 19, 116–126.
6. Araghinejad, S. (2014). Data-driven modeling: using MATLAB in water resources and environmental engineering. Journal of Water Science and Technology Library, 67, 265.
7. Belayneh, A., & Adamowski, J. (2012). Standard precipitation index drought forecasting using neural networks, wavelet neural networks, and support vector regression. Journal of Applied Computational Intelligence and Soft Computing, 1–13. https://doi.org/10.1155/2012/794061
8. Cannas, B., Alessandra, F., See, L., & Sias, G. (2006). Data preprocessing for river flow forecasting using neural networks: Wavelet transforms and data partitioning. Physics and Chemistry of the Earth, 31(18), 1164–1171. https://doi.org/1016/j.pce.2006.03.020
9. Chakraborty, S., & Biswas, S. (2023). River discharge prediction using wavelet-based artificial neural network and long short-term memory models: a case study of Teesta River Basin, India. Stochastic Environmental Research and Risk Assessment, 37(8), 1-22. https://doi.org/1007/s00477-023-02443-y
10. Chen, C.H. (1999). Wevelet approach to optimizing dynamic system. Control Theory and Applications. IEE Proceedings, 146(2), 213-219. https://doi.org/1049/ip-cta:19990516
11. Danandeh Mehr, A., Kahya, E., & Olyaie, E. (2013). Streamfow prediction using linear genetic programming in comparison with a neurowavelet technique. Journal of Hydrology, 505, 240–249. https://doi.org/1016/j.jhydrol. 2013.10.003
12. Dastorani, M.T., Hajibigloo, M., & Shojaee, H. (2022). Identification of the land use changes on river flooding bed, affective on reservoir water quality (Case study: headwater of Kardeh reservoir). Geography and Development, 20(66), 255-282. https://doi.org/22111/J10.22111.2022.6739
13. Dalkiliç, H.Y., & Hashimi, S.A. (2020). Prediction of daily streamfow using artificial neural networks (ANNs), wavelet neural networks (WNNs), and adaptive neuro-fuzzy inference system (ANFIS) models. Water Supply, 20(4), 1396–1408. https://doi.org/2166/ws.2020.062
14. Freire, P.K.d.M.M., Santos, C.A.G., & da Silva, G.B.L. (2019). Analysis of the use of discrete wavelet transforms coupled with ANN for shortterm streamflow forecasting. Apply Soft Computer, 80, 494–505. https://doi.org/1016/ j.asoc.2019.04.024
15. Graps A. (1995). An Instroduction to wavelet. Computing in Science and Engineering, 2, 50-61. https://doi.org/ 1109/99.388960
16. Grossmann, A., & Morlet, J. (1984). Decomposition of Hardy functions into square integrable wavelets of constant shape. SIAM Journal on Mathematical Analysis, 15(4), 723–736. https://doi.org/1137/0515056
17. Güneş, M., Parim, C., Yıldız, D., & Büyüklü, A. (2021). Predicting monthly streamflow using a hybrid wavelet neural network: case study of the Çoruh River Basin. Polish Journal of Environmental Studies, 30(4), 3065–3075. https://doi.org/15244/pjoes/130767
18. Jimeno-Saez, P., Senent-Aparicio, J., Perez-Sanchez, J., & PulidoVelazquez, D. (2018). A comparison of SWAT and ANN models for daily runof simulation in diferent climatic zones of peninsular Spain. Water, 10(2), 192. https://doi.org/3390/w10020192
19. Katipoglu, O.M. (2023). Monthly streamfow prediction in Amasya, Türkiye, using an integrated approach of a feedforward backpropagation neural network and discrete wavelet transform. Modeling Earth Systems and Environment, 9(2), 1-13. https://doi.org/1007/s40808-022-01629-7
20. Kim, T.W., Valdes, J.B. (2003). Nonlinear model for drought forecasting based on a conjunction of wavelet transforms and neural networks. Journal of Hydrologic Engineering, 8(6), 319–328. https://doi.org/1061/(ASCE) 1084-0699(2003)8:6(319)
21. Maheswaran, R., Khosa, R. (2012). Comparative study of different wavelets for hydrologic forecasting. Computers and Geosciences, 46, 284–295. https://doi.org/1016/j.cageo.2011.12.015
22. Mallat, S. (1999). A Wavelet Tour of Signal Processing, 2nd ed. Academic, New York.
23. Misiti, M., Misiti, Y., Oppenheim, G., & Poggi, J.M. (2001). Wvelet toolbox for use with Matlab.
24. Modaresi, F., Araghinejad, S., & Ebrahimi, K. (2017). A comparative assessment of artificial neural network, generalized regression neural network, least-square support vector regression, and k-nearest neighbor regression for monthly streamflow forecasting in linear and nonlinear conditions. Water Resources Management, 32(5), 243-258. https://doi.org/1007/s11269-017-1807-2
25. Momeneh, S., & Nourani, V. (2022). Application of a novel technique of the multi‑discrete wavelet transforms in hybrid with artifcial neural network to forecast the daily and monthly streamfow. Modeling Earth Systems and Environment, 8(2), 4629–4648. https://doi.org/1007/s40808-022-01387-6
26. Moriasi, D.N., Arnold, J.G., & Van Liew, M.W. (2007). Model evaluation guidelines for systemic quantification of accuracy in watershed simulations. Transactions of ASABE, 50(3), 885-900. https://doi.org/13031/2013.23153
27. Nayak, P., Venkatesh, B., Krishna, B., & Jain, S.K. (2013). Rainfall-runof modeling using conceptual, data driven, and wavelet based computing approach. Journal of Hydrology, 493, 57–67. https://doi.org/1016/j.jhydrol. 2013.04.016
28. Okkan, U. (2013). Wavelet neural network model for reservoir inflow prediction. Scientia Iranica, 19, 1445-1455. https://doi.org/10.1016/j.scient.2012.10.009
29. Partal, T., Kisi, O. (2007). Wavelet and neuro-fuzzy conjunction model for precipitation forecasting. Journal of Hydrology 342(1):199–212. https://doi.org/1016/j.jhydrol.2007.05.026
30. Poonia, V., & Tiwari, H.L. (2020). Rainfall-runoff modeling for the Hoshangabad Basin of Narmada River using artificial neural network. Arabian Journal of Geosciences, 13(18), 1–10. https://doi.org/1007/s12517-020-05930-6
31. Rajendra, P., Murthy, K.V.N., Subbarao, A., & Boadh, R. (2019). Use of ANN models in the prediction of meteorological data. Modeling Earth Systems and Environment, 5(14), 1051–1058. https://doi.org/1007/s40808-019-00590-2
32. Santos, C.A.G., & Silva, G.B.L. (2014). Daily streamflow forecasting using a wavelet transform and artificial neural network hybrid models. Hydrological Sciences Journal 59(2) :312–324. https://doi.org/1080/02626667.2013.800944
33. Santos, C.A.G., Freire, P.K.M.M., Silva, G.B.L., & Silva, R.M. (2014). Discrete wavelet transform coupled with ANN for daily discharge forecasting into Três Marias reservoir. Proceedings of the International Association of Hydrological Sciences, 364, 100–105. https://doi.org/10.5194/piahs-364-100-2014
34. Shoaib, M., Shamseldin, A.Y., & Melville, B.W. (2014). Comparative study of different wavelet based neural network models for rainfall-runoff modelling. Journal of Hydrology, 515(1-2), 47-58. https://doi.org/1016/ j.jhydrol.2014.04.055
35. Siddiqi, T.A., Ashraf, S., Khan, S.A., & Iqbal, M.J. (2021). Estimation of datadriven streamfow predicting models using machine learning methods. Arabian Journal of Geosciences, 14(11), 1058-1567. https://doi.org/1007/ s12517-021-07446-z
36. Sithara, S., Pramada, S.K., & Thampi, S.G. (2020). Sea level prediction using climatic variables: a comparative study of SVM and hybrid wavelet SVM approaches. Acta Geophysica, 68, 1779–1790. https://doi.org/1007/s11600-020-00484-3
37. Sun, Y., Niu, J., & Sivakumar, B. (2019). A comparative study of models for short-term streamflow forecasting with emphasis on waveletbased approach. Stochastic Environmental Research and Risk Assessment, 33, 1875–1891. https://doi.org/1007/s00477-019-01734-7
38. Tareke, K.A., & Awoke, A.G. (2023). Hydrological drought forecasting and monitoring system development using artificial neural network (ANN) in Ethiopia. Heliyon, 9(2). https://doi.org/1016/j.heliyon.2023.e13287
39. Tayyab, M., Zhou, J., Dong, X., Ahmad, I., & Sun, N. (2019). Rainfall-runof modeling at Jinsha River basin by integrated neural network with discrete wavelet transform. Meteorology Atmospheric Physics, 131(1), 115–125. https://doi.org/1007/s00703-017-0546-5
40. Tiwari, D.K., Tiwari, H.L., & Nateriya, R. (2022). Runoff modeling in Kolar river basin using hybrid approach of wavelet with artificial neural network. Journal of Water and Climate Change, 13(3), 963. https://doi.org/2166/ wcc.2021.246
41. Tiwari, M.K., & Chatterjee, C. (2011). A new wavelet-bootstrap-ANN hybrid model for daily discharge forecasting. Journal of Hydroinformatics, 13(3), 500–519. https://doi.org/2166/hydro.2010.142
42. Wagena, M.B., Goering, D., Collick, A.S., Bock, E., Fuka, D.R., Buda, A., & Easton, Z.M. (2020). Comparison of short-term streamfow forecasting using stochastic time series, neural networks, process-based, and Bayesian models. Environmental Modelling & Software, 126(4), 104669. https://doi.org/1016/j.envsoft.2020.104669
43. Wambua, R.M. (2014). Drought forecasting using indices and artificial neural networks for upper tana River Basin, Kenya-A review concept. Journal of Civil & Environmental Engineering, 4(4), 1-12. https://doi.org/4172/2165-784X.1000152
44. Wang, W., & Ding, J. (2003). Wavelet network model and its application to the prediction of hydrology. Nature and Science, 1(1), 67-71.
45. Yilmaz, M., Tosunoğlu, F., Kaplan, N.H., Üneş, F., & Hanay, Y.S. (2022). Predicting monthly streamfow using artifcial neural networks and wavelet neural networks models. Modeling Earth Systems and Environment, 8(4), 3-20. https://doi.org/1007/s40808-022-01403-9
46. Young, C.C., Liu, W.C., & Wu, M.C. (2017). A physically based and machine learning hybrid approach for accurate rainfallmodeling during extreme typhoon events. Applied Soft Computing, 53(3-4), 205–216. https://doi.org/1016/ j.asoc.2016.12.052