JBT Journal
Advanced

Raqamli Transformatsiya va Sun’iy Intellekt

TRANSPORT-LOGISTIKA TIZIMIDA YUK OQIMINI PROGNOZLASH UCHUN LSTM VA GRU MODELLARINI TAQQOSLASH.

Abdumalikov Akmaljon Abduxoliq o‘g‘li1, Irgasheva Umida1

1 Mirzo Ulug‘bek nomidagi O‘zbekiston Milliy universitetining Jizzax filiali

Zenodo DOI: 10.5281/zenodo.23255142

Keywords

YOLO prognozlash LSTM chuqur o‘qitish vaqt qatori MAPE RMSE CCTV transport-logistika tizimi R² MAE yuk oqimi adaptiv marshrut boshqarish. GRU

Abstract

Ushbu maqolada transport-logistika tizimida yuk oqimini vaqt qatori sifatida prognozlash uchun LSTM (Long Short-Term Memory) va GRU (Gated Recurrent Unit) neyron tarmoq modellarini taqqoslash va baholash masalasi ko‘rib chiqiladi. Tadqiqotda transport oqimining vaqt bo‘yicha o‘zgarishini ifodalovchi ma’lumotlardan foydalanish, ularni oldindan qayta ishlash, o‘qitish va prognozlash bosqichlari tavsiflanadi. LSTM uzoq muddatli bog‘lanishlarni xotira katagi va uchta asosiy darvoza orqali saqlasa, GRU yangilash va qayta tiklash darvozalari yordamida nisbatan sodda arxitekturani ta’minlaydi. Modellarni baholashda MAE, RMSE, MAPE va R² ko‘rsatkichlari, shuningdek, o‘qitish vaqti, parametrlar soni hamda hisoblash resurslariga bo‘lgan talablar hisobga olinadi. Maqolada CCTV va YOLO asosida transport vositalarini aniqlash va sanashdan olingan ma’lumotlarni LSTM/GRU prognozlash modellariga uzatish imkoniyati ham ko‘rsatiladi. Bunday yondashuv keyingi bosqichda adaptiv marshrut boshqarish algoritmlarini ma’lumot bilan ta’minlashga xizmat qiladi. LSTM yoki GRU ning ustunligi oldindan qat’iy belgilanmaydi, u ma’lumotlar to‘plami, prognoz gorizonti, belgilar tarkibi va tajriba sharoitlariga bog‘liq holda aniqlanishi kerak.

Views: 100 Published: 2026-10-09 Pages: 47-52 Volume: 4 Issue: 5

References

  1. A comprehensive review of traffic flow forecasting based on deep learning. (2026). Neurocomputing, 668, 132269. https://doi.org/10.1016/j.neucom.2025.132269
  2. Incorporating multivariate auxiliary information for traffic prediction on highways. (2023). Sensors, 23(7), 3631. https://doi.org/10.3390/s23073631
  3. Irgasheva U. A., Ergashev A. A. & Bo‘riboyev A. (2026). Sun’iy intellekt va big data asosida fan va ta’lim integratsiyasining ilmiy-nazariy asoslari. In Raqamli jamiyatda fan va ta’limning integratsiyalashuvi (pp. 267–270).
  4. Jang, H.-C., & Chen, C.-A. (2024). Urban traffic flow prediction using LSTM and GRU. Engineering Proceedings, 55(1), 86. https://doi.org/10.3390/engproc2023055086
  5. Abdumalikov A.A. & Irgasheva U.A. (2026) FACENET yordamida belgilarni avtomatik o‘rganish orqali yuzning indivdual xususiyatlarini aniqlash. In Al-Farg‘oniy avlodlari. (pp. 98-102) https://al-fargoniy.uz/index.php/journal/article/view/1240/2928
  6. Reza, S., Ferreira, M. C., Machado, J. J. M., & Tavares, J. M. R. S. (2022). A multi-head attention-based transformer model for traffic flow forecasting with a comparative analysis to recurrent neural networks. Expert Systems with Applications, 202, 117275. https://doi.org/10.1016/j.eswa.2022.117275
  7. Applying hybrid LSTM-GRU model based on heterogeneous data sources for traffic speed prediction in urban areas. (2022). Sensors, 22(9), 3348. https://doi.org/10.3390/s22093348
  8. Performance analysis of neural network architectures for time series forecasting: A comparative study of RNN, LSTM, GRU, and hybrid models. (2025). MethodsX, 15, 103462. https://doi.org/10.1016/j.mex.2025.103462
  9. Irgasheva U. A. (2026). CNN arxitekturalari yordamida belgilarni avtomatik o‘rganish orqali yuzning individual xususiyatlarini aniqlash. In Yangi O‘zbekiston taraqqiyotida yosh olimlarning o‘rni (pp. 582–585).