BETONLASHTIRILMAGAN MAGISTRAL KANALLARDA TUPROQQA SHIMILISH YO‘QOTISHLARINI HISOBGA OLUVCHI GIBRID SAINT-VENANT VA LSTM NEYRON TARMOG‘I MODELI
Keywords:
Kalit so‘zlar: Betonlashtirilmagan kanallar, Saint-Venant tenglamalari, LSTM neyron tarmog‘i, Fizika bilan cheklangan chuqur o‘qitish (PINN), Tuproqqa shimilish yo‘qotishlari, Raqamli egizak, Preissmann sxemasiAbstract
Annotatsiya: Global iqlim o‘zgarishi va chuchuk suv resurslari taqchilligi sharoitida, betonlashtirilmagan (tabiiy tuproq o‘zanli) magistral kanallarda translyatsiya jarayonidagi dynamik filtratsiya yo‘qotishlarini onlayn hisoblash va suv taqsimotini optimallashtirish dolzarb muandislik vazifasi hisoblanadi. Mazkur tadqiqotda qo‘lda boshqariladigan shlyuz inshootlariga ega magistral kanal misolida fizik qonuniyatlar bilan cheklangan gibrid Saint-Venant va Long Short-Term Memory (LSTM) neyron tarmog‘i modeli taklif etilgan. Ob’ektda metrologik datchiklarning mutloq tanqisligi muammosi modifikatsiyalangan bir o‘lchamli Saint-Venant tenglamalari va Preissmann implicit raqamli boks sxemasi solveri poydevorida kanalning virtual “Raqamli Egizaki”ni qurish hamda 35 040 ta vaqt qadamidan iborat yuqori aniqlikdagi sintetik dataset generatsiya qilish orqali hal etildi. Mustaqil test ma’lumotlarida olingan natijalar gibrid modelning suv sathini bashorat qilishda o‘ta yuqori aniqlikka (RMSE = 0.024 m, R2 = 0.975) ega ekanligini va bitta optimallash sikli vaqtini 1.8 sekundgacha qisqartirib, real vaqt rejimida qaror qabul qilish imkonini ko‘rsatdi.
References
1. Seytov A., Abduraxmonov O., Choriyorov N., Maxkamov А. Development of algorithms for planning and management of water resources in Irrigation systems in the conditions of climate chаnging, SCIENTIFIC BULLETIN Physical and Mathematical Research 2024/№1(6).
2. Zhang, D., & Wang, Z. A hybrid physics-informed LSTM network for unsteady open-channel flow routing and water level prediction. Journal of Hydrology, 618, (2023), 129204.
3. Li, S., Wu, H., & Cheng, L. Numerical stability and discretization schemes for shallow water equations using automated simulator engines. Advances in Water Resources, 172, (2024), 104512.
4. Al-Musawi, W. M., & Al-Suhaili, S. B. Dynamic estimation of seepage losses in unlined earthen irrigation canals under varying hydrostatic pressure. Irrigation and Drainage, 71(4), (2022), 512-524.
5. Goodfellow, I., Bengio, Y., & Courville, A. Deep Learning. MIT Press(2016).
6. Wang, J., & Raza, A. Synthetic dataset generation methodologies for data-scarce hydraulic systems using numerical twins. Water Resources Management, 39(2), (2025), 345-361.
7. Krause, S., Boyle, D. P., & Base, F. Comparison of different efficiency criteria for hydrological model assessment. Advances in Geosciences, 5, (2005), 89-97.
8. Liu, T., Feng, X., & Zhou, Y. Computationally efficient real-time control of large-scale irrigation canals using deep neural network surrogates. Agricultural Water Management, 284, (2023), 108341.
9. Clemmens, A. J., & Schuurmans, J. Progress in open-channel irrigation canal control. Journal of Irrigation and Drainage Engineering, 130(1), (2004), 25-34.
10. Babaeian, M., & Rahman, M. M. Revisiting the Saint-Venant equations for transient open-channel flows using modern mathematical solver architectures. Journal of Hydraulic Engineering, 149(3), (2023), 04022031
11. Khosravi, A., & Han, D. Numerical stability and boundary condition tracking of the four-point implicit Preissmann scheme in automated canal simulation software. Advances in Water Resources, 161, (2022), 104118.
12. An, J., Zhang, L., & Li, X. Geometric effects on hydraulic radius and wetting perimeter in trapezoidal earthen canals under structural deformation. Agricultural Water Management, 291, (2025), 108560.
13. Turaev R., Seytov A., Haydarova R., Abduraxmonov O., Choriyorov N. “Optimal water management in the channels of machine water raise systems”, O‘zMU xabarlari 21.1, 2024.
14. Karniadakis, G. E., Kevrekidis, I. G., Lu, L., Perdikaris, P., Wang, S., & Yang, L. Physics-informed machine learning. Nature Reviews Physics, 3(6), (2021), 422-440.
