O‘zbekiston qishloq xo‘jaligi Volume 2026 Issue 5 (2026) · pp. 46-48
SUV RESURSLARINI MONITORING QILISH VA BASHORATLASHNING INTELLEKTUAL MULTISENSOR TIZIMI METODLARI VA ALGORITMLARI
Ozodov, Ezozbek
Abstract
This paper addresses the challenge of multisensor data integration and intelligent forecasting for water resources management in regions experiencing critical water scarcity, exemplified by the Aydarkul Lake system in Uzbekistan which decreased from 44 to 10 billion m³ (2000-2015), resulting in $115 million USD economic losses. We propose an intelligent multisensor measuring system based on hybrid machine learning models (Random Forest, Gradient Boosting, LSTM) integrating real-time data from TDS, pH, ultrasonic level, flow, and pressure sensors. The framework achieves 95% accuracy for short-term (1-7 days) and 85% for long-term (1 year) forecasts, reducing water losses by 23% and improving irrigation reliability by 18% compared to conventional approaches.
Multisensor system, intelligent forecasting, machine learning, water resources, Aydarkul Lake, real-time monitoring, LSTMМультисенсорная система, интеллектуальное прогнозирование, машинное обуче- ние, водные ресурсы, озеро Айдаркуль, мониторинг в реальном времени, LSTMMultisensor tizim, intellektual bashoratlash, mashinaviy o’qitish, suv resurslari, Aydarko’l, real vaqt monitoringi, LSTM
Metadata source: the journal's OAI-PMH archive · Sindex does not store the full text; it links to the source.