مقايسه تأثير هندسه کانالهاي مستطيلي و ذوزنقهاي بر ضريب دبي سرريزهاي جانبي مستطيلي با استفاده از مدلهاي يادگيري ماشين بهينهشده
محورهای موضوعی : کاربرد کامپیوتر در مسائل آب و خاک
کیومرث روشنگر
1
*
,
آیدین پناهی
2
1 - استاد، گروه مهندسي منابع آب، دانشکده مهندسي عمران، دانشگاه تبريز، تبريز، ايران.
2 - دانشجو دکتري، گروه مهندسي منابع آب، دانشکده مهندسي عمران، دانشگاه تبريز، تبريز، ايران.
کلید واژه: الگوريتمهاي فرا ابتکاري, آناليز حساسيت, سرريز جانبي, ضريب دبي, يادگيري ماشين,
چکیده مقاله :
زمينه و هدف: سرريزهاي جانبي نقشي کليدي در اندازهگيري و کنترل جريان در سامانههاي هيدروليکي ايفا ميکنند، بهويژه در پروژههاي کاهش خسارات سيلاب و مديريت رواناب. برآورد دقيق ضريب دبي (Cd) اين سازهها، به دليل تعامل پيچيده بين ديناميک جريان و هندسه کانال، به يک چالش فني مهم تبديلشده است. در اين پژوهش، براي نخستين بار، دو مدل ترکيبي نوين تحت عنوان SVM-HOA و SVM-RSA معرفيشدهاند که در آنها، ماشين بردار پشتيبان (SVM) به ترتيب با الگوريتمهاي بهينهسازي اسب (HOA) و جستوجوي خزنده (RSA) تنظيمشده است. اين مدلها بهمنظور پيشبيني ضريب دبي متناسب با هندسه براي سرريزهاي جانبي مستطيلي لبهتيز در دو نوع کانال مستطيلي و ذوزنقهاي به کار گرفتهشدهاند.
روش پژوهش: اين مطالعه، دو سناريوي آزمايشگاهي بررسيشده است. سناريوي نخست شامل 407 داده آزمايشگاهي از سرريزهاي جانبي مستطيلي لبهتيز در کانالهاي مستطيلي و سناريوي دوم شامل 78 داده آزمايشگاهي در کانالهاي ذوزنقهاي است. در ابتدا، مهمترين پارامترهاي بيبعد مؤثر ازجمله Fr1، P/y1، L/B، L/y1، y1/B و m شناسايي شدند. اين پارامترها بهعنوان ورودي مدلهاي SVM استفاده شدند. هر مدل ترکيبي با بهينهسازي ابر پارامترهاي SVM بهوسيله الگوريتمهاي HOA و RSA پيادهسازي شد تا از افتادن در کمينههاي موضعي جلوگيري شود و دقت پيشبيني افزايش يابد. دادهها به نسبت 80 به 20 براي آموزش و آزمون تقسيم شدند و براي جلوگيري از بيش برازش، از اعتبارسنجي متقابل 10 بخشي استفاده شد. عملکرد مدلها با شاخصهاي آماري استاندارد ازجمله خطاي ريشه ميانگين مربعات (RMSE)، کارايي نش-ساتکليف (NSE) و ضريب همبستگي (R) ارزيابي گرديد. همچنين، تحليل حساسيت براي بررسي ميزان تأثير هر پارامتر ورودي بر خروجي انجام شد.
نتايج و بحث: نتايج ارزيابي تطبيقي نشان داد که هر دو مدل SVM-HOA و SVM-RSA در پيشبيني دقيق ضريب دبي در هر دو نوع کانال عملکرد مناسبي داشتند. در سناريوي اول (کانال مستطيلي)، مدل SVM-HOA با وروديهاي نظير نسبت ارتفاع سرريز به عمق جريان ابتدايي (P/y1)، نسبت طول تاج سرريز به عرض کف کانال (L/B)، نسبت طول تاج به عمق جريان (L/y1) و عدد فرود بالادست (Fr1) مقادير NSE، RMSE و R به ترتيب برابر با 0.904، 0.063 و 0.953 بهترين عملکرد را نشان داد. در سناريوي دوم (کانال ذوزنقهاي) نيز مدل SVM-HOA با وروديهاي L/B، Fr1، y1/B و شيب ديواره کانال (m) به مقادير NSE، RMSE و R به ترتيب برابر با 0.964، 0.020 و 0.986 دستيافت. نمودارهاي تيلور، برتري مدل SVM-HOA نسبت به SVM-RSA را در قدرت پيشبيني و پايداري نتايج تأييد کردند. تحليل حساسيت نشان داد که نسبت P/y1 در کانال مستطيلي و عدد فرود بالادست (Fr1) در کانال ذوزنقهاي بيشترين تأثير را بر مقدار Cd دارند. اين يافتهها بر اهميت توجه به متغيرهاي هندسه محور و انتخاب الگوريتم بهينهسازي مناسب در کاربردهاي هيدروليکي تأکيد ميکنند.
نتيجهگيري: اين پژوهش رويکردي نوآورانه و متکي بر هندسه را براي پيشبيني دقيق ضريب دبي در سرريزهاي جانبي مستطيلي لبهتيز ارائه ميدهد. دو مدل ترکيبي پيشنهادي، يعني SVM-HOA و SVM-RSA، دقت بالا، تطبيقپذيري مناسب با هندسههاي مختلف کانال و مقاومت بالا در برابر بيش برازش و کمينههاي محلي را از خود نشان دادند. نتايج بيانگر برتري قابلتوجه مدل SVM-HOA در بيشتر موارد است، بهويژه زماني که با پارامترهاي بيبعد مناسب و نسبت بهينهي دادههاي آموزش و تست همراه باشد. اين مدلها ميتوانند بهعنوان ابزاري کارآمد براي مهندسان هيدروليک در طراحي، شبيهسازي و تحليل عملکرد سرريزهاي جانبي در شرايط مختلف جريان و انواع کانالها مورداستفاده قرار گيرند.
Introduction: Side weirs play a key role in flow measurement and control in hydraulic systems, especially in flood mitigation and runoff management projects. Accurate estimation of the discharge coefficient (Cd) has become a critical technical challenge due to the complex interactions between flow dynamics and channel geometry. This study introduces, for the first time, two novel hybrid models—SVM-HOA and SVM-RSA—based on SVM optimized by the Horse Optimization Algorithm and Reptile Search Algorithm, respectively. These models are applied to predict the geometry-specific discharge coefficient of sharp-crested rectangular side weirs in both rectangular and trapezoidal channel configurations.
Method: The proposed methodology involves two experimental scenarios. Scenario 1 includes 407 laboratory observations of sharp-crested rectangular side weirs located in rectangular channels, whereas Scenario 2 consists of 78 observations in trapezoidal channels. Dimensional analysis was first conducted to identify the most influential nondimensional parameters, including Fr1, P/y1, L/B, L/y1, y1/B, and m. These parameters served as inputs for training the SVM models. Each hybrid model was implemented by optimizing the SVM hyperparameters using the HOA and RSA metaheuristic algorithms, aiming to avoid local minima and enhance prediction accuracy. The dataset was split into training and testing subsets in a ratio: 80:20. To mitigate overfitting, 10-fold cross-validation was applied. Model performance was evaluated using standard statistical metrics: Root Mean Square Error (RMSE), Nash–Sutcliffe Efficiency (NSE), and correlation coefficient (R). Additionally, sensitivity analysis was conducted to assess the influence of each input variable on the final prediction.
Results: Results of the comparative evaluation showed that both SVM-HOA and SVM-RSA yielded high accuracy in predicting Cd for both channel configurations. In Scenario 1 (rectangular channel), With inputs such as the ratio of the spillway height to the initial flow depth (P/y1), the ratio of the spillway crest length to the channel bottom width (L/B), the ratio of the crest length to the flow depth (L/y1), and the upstream Froude number, the best-performing model was SVM-HOA with R = 0.953, RMSE = 0.063, and NSE = 0.904 during the testing phase. In Scenario 2 (trapezoidal channel), the optimal model also used SVM-HOA With inputs L/B, Fr1, y1/B and channel wall slope (m) and achieved R = 0.986, RMSE = 0.020, and NSE = 0.964. Taylor diagrams confirmed the robustness and higher predictive power of the SVM-HOA model compared to SVM-RSA. Sensitivity analysis revealed that in the rectangular channel setup, the P/y1 ratio was the most influential input, while in the trapezoidal channel, the upstream Froude number (Fr1) played the dominant role in determining Cd. The findings underscore the importance of accounting for geometry-specific variables and selecting appropriate optimization algorithms to fine-tune ML models for hydraulic applications.
Conclusions: This study presents a novel and geometry-aware approach to accurately predicting the discharge coefficient of sharp-crested rectangular side weirs using two advanced hybrid SVM-based models. The proposed SVM-HOA and SVM-RSA frameworks demonstrated consistently high accuracy, excellent adaptability to different channel geometries, and strong robustness against overfitting and local minima. Results indicate that SVM-HOA significantly outperforms SVM-RSA in most cases, particularly when paired with well-chosen nondimensional parameters and a properly tuned training-testing data split. The models developed in this work can effectively assist hydraulic engineers in the design, simulation, and performance analysis of side weirs under diverse flow conditions and channel types.
Abdollahi, A., Kabiri-Samani, A., Asghari, K., Atoof, H., & Bagheri, S. (2017). Numerical modeling of flow field around the labyrinth side-weirs in the presence of guide vanes. ISH Journal of Hydraulic Engineering, 23(1), 71–79. https://doi.org/10.1080/09715010.2016.1239555
Abualigah, L., Abd Elaziz, M., Sumari, P., Geem, Z. W., & Gandomi, A. H. (2022). Reptile Search Algorithm (RSA): A nature-inspired meta-heuristic optimizer. Expert Systems with Applications, 191, 116158. https://doi.org/10.1016/j.eswa.2021.116158
Agrawal, S. K. (1989). Elementary Discharge Coefficient of Rectangular Side Weirs. Roorkee Univ Roorkee, India.
Azamathulla, H. M., Haghiabi, A. H., & Parsaie, A. (2016). Prediction of side weir discharge coefficient by support vector machine technique. Water Science and Technology: Water Supply, 16(4), 1002–1016. https://doi.org/10.2166/ws.2016.014
Azimi, H., Bonakdari, H., & Ebtehaj, I. (2017). Sensitivity analysis of the factors affecting the discharge capacity of side weirs in trapezoidal channels using extreme learning machines. Flow Measurement and Instrumentation, 54, 216–223. https://doi.org/10.1016/j.flowmeasinst.2017.02.005
Azimi, H., Shabanlou, S., & Salimi, M. S. (2014). Free surface and velocity field in a circular channel along the side weir in supercritical flow conditions. Flow Measurement and Instrumentation, 38, 108–115. https://doi.org/10.1016/j.flowmeasinst.2014.05.013
Bagheri, S., Kabiri-Samani, A. R., & Heidarpour, M. (2014). Discharge coefficient of rectangular sharp-crested side weirs Part II: Domínguez’s method. Flow Measurement and Instrumentation, 35, 116–121. https://doi.org/10.1016/j.flowmeasinst.2013.10.006
Balahang, S., & Ghodsian, M. (2024). Enhancing rectangular side weir discharge prediction using stacking technique. Flow Measurement and Instrumentation, 97, 102622. https://doi.org/10.1016/j.flowmeasinst.2024.102622
Biggiero, V., Longobardi, D., & Pianese, D. (1994). Indagine sperimentale su sfioratori laterali a soglia bassa. Giornale Del Genio Civile, 132(7-8–9), 183–199.
Bilhan, O., Emiroglu, M. E., & Kisi, O. (2010). Application of two different neural network techniques to lateral outflow over rectangular side weirs located on a straight channel. Advances in Engineering Software, 41(6), 831–837. https://doi.org/10.1016/j.advengsoft.2010.03.001
Borghei, S. M., Jalili, M. R., & Ghodsian, M. (1999). Discharge coefficient for sharp-crested side weir in subcritical flow. Journal of Hydraulic Engineering, 125(10), 1051–1056. https://doi.org/10.1061/(ASCE)0733-9429(1999)125:10(1051)
Cheong, H.-F. (1991). Discharge coefficient of lateral diversion from trapezoidal channel. Journal of Irrigation and Drainage Engineering, 117(4), 461–475. https://doi.org/10.1061/(ASCE)0733-9437(1991)117:4(461)
Cortes, C., & Vapnik, V. (1995). Support-vector networks. Machine Learning, 20, 273–297. https://doi.org/10.1007/Bf00994018
Dhar, M., Sapkota, S. C., Saha, P., & Arora, S. (2025). Evaluating Discharge Coefficient of Rectangular Sharp Crested Weirs Using Machine Learning Models. Water Resources Management, 39(8), 4151-4171. https://doi.org/10.1007/s11269-025-04152-3
Ebtehaj, I., Bonakdari, H., Zaji, A. H., Azimi, H., & Khoshbin, F. (2015). GMDH-type neural network approach for modeling the discharge coefficient of rectangular sharp-crested side weirs. Engineering Science and Technology, an International Journal, 18(4), 746–757. https://doi.org/10.1016/j.jestch.2015.04.012
El-Khashab, A., & Smith, K. V. H. (1976). Experimental investigation of flow over side weirs. Journal of the Hydraulics Division, 102(9), 1255–1268. https://doi.org/10.1061/JYCEAJ.0004610
Emiroglu, M. E., Agaccioglu, H., & Kaya, N. (2011). Discharging capacity of rectangular side weirs in straight open channels. Flow Measurement and Instrumentation, 22(4), 319–330. https://doi.org/10.1016/j.flowmeasinst.2011.04.003
Ghaderi, A., Dasineh, M., Abbasi, S., & Abraham, J. (2020). Investigation of trapezoidal sharp-crested side weir discharge coefficients under subcritical flow regimes using CFD. Applied Water Science, 10(1), 1–12. https://doi.org/10.1007/s13201-019-1112-8
Granata, F., Gargano, R., & Santopietro, S. (2016). A flow field characterization in a circular channel along a side weir. Flow Measurement and Instrumentation, 52, 92–100. https://doi.org/10.1016/j.flowmeasinst.2016.09.011
Hager, W. H. (1994). Supercritical Flow in Circular‐Shaped Side Weir. Journal of Irrigation and Drainage Engineering, 120(1), 1–12. https://doi.org/10.1061/(ASCE)0733-9437(1994)120:1(1)
Hien, L. T. T., Van Chien, N., & Xuan-Hien, L. (2024). Advanced hybrid techniques for predicting discharge coefficients in ogee-crested spillways: integrating physical, numerical, and machine learning models. Environmental Research Communications, 6(11), 115002. https://doi.org/10.1088/2515-7620/ad8a24
Jalili, M. R., & Borghei, S. M. (1996). Discussion: discharge coefficient of rectangular side weirs. Journal of Irrigation and Drainage Engineering, 122(2), 132. https://doi.org/10.1061/(ASCE)0733-9437(1996)122:2(132)
Khoshbin, F., Bonakdari, H., Ashraf Talesh, S. H., Ebtehaj, I., Zaji, A. H., & Azimi, H. (2016). Adaptive neuro-fuzzy inference system multi-objective optimization using the genetic algorithm/singular value decomposition method for modelling the discharge coefficient in rectangular sharp-crested side weirs. Engineering Optimization, 48(6), 933–948. https://doi.org/10.1080/0305215X.2015.1071807
MiarNaeimi, F., Azizyan, G., & Rashki, M. (2021). Horse herd optimization algorithm: A nature-inspired algorithm for high-dimensional optimization problems. Knowledge-Based Systems, 213, 106711. https://doi.org/10.1016/j.knosys.2020.106711
Raju, K. G. R., Gupta, S. K., & Prasad, B. (1979). Side Weir in Rectangular Channel. Journal of the Hydraulics Division, 105(5), 547–554. https://doi.org/10.1061/JYCEAJ.0005207
Roushangar, K., Khoshkanar, R., & Shiri, J. (2016). Predicting trapezoidal and rectangular side weirs discharge coefficient using machine learning methods. ISH Journal of Hydraulic Engineering, 22(3), 254–261. https://doi.org/10.1080/09715010.2016.1177740
Roushangar, K., Saadatjoo, R., Abbaszadeh, H., & Panahi, A. (2024a). Estimation of air concentration in chute spillway using metamodel methods. Iranian Journal of Soil and Water Research, 55(4), 601–613. DOI: https://doi.org/10.22059/ijswr.2024.370643.669640 [in Persian].
Roushangar, K. , Abbaszadeh, H. , Saadatjoo, R. and Panahi, A. (2024b). Prediction of Air Concentration in Stepped Spillways Using Data-Oriented Methods. Environment and Water Engineering, 11(3), 316-325. DOI: https://doi.org/10.22034/ewe.2024.433409.1905
Saffar, S., Babarsad, M. S., Shooshtari, M. M., & Riazi, R. (2021). Prediction of the discharge of side weir in the converge channels using artificial neural networks. Flow Measurement and Instrumentation, 78, 101889. https://doi.org/10.1016/j.flowmeasinst.2021.101889
Seyedian, S. M., & Kisi, O. (2024). Uncertainty analysis of discharge coefficient predicted for rectangular side weir using machine learning methods. Journal of Hydrology and Hydromechanics, 72(1), 113-130. https://doi.org/10.2478/johh-2023-0043
Subramanya, K., & Awasthy, S. C. (1972). Spatially varied flow over side-weirs. Journal of the Hydraulics Division, 98(1), 1–10. https://doi.org/10.1061/JYCEAJ.0003188