Predicció avançada de la humitat del terreny en camps valencians d'Agricultura de Precisió amb optimització per a IoT
Year: 2026
Type: dissertation
Abstract: [EN] The integration of artificial intelligence into precision agriculture represents a transformative advancement in modern farming, essential for optimizing labor efficiency and maximizing field productivity. This study addresses the critical domain of soil moisture prediction within irrigation systems for irrigated agricultural fields. This study develops an advanced system for soil moisture prediction in agricultural fields within the Valencian Community, integrating in-situ sensors, Precision Agriculture techniques, and machine learning models optimized for IoT devices. The research utilizes time-series data of soil moisture at various depths, alongside agro-climatic variables (temperature, precipitation, radiation, and rainfall) collected from IoT nodes in representative citrus crop plots. Various predictive algorithms—ranging from traditional Machine Learning (ML) to Recurrent Neural Networks (RNN)—are evaluated, comparing their performance metrics and inference times. The results establish a replicable procedure for forecasting soil moisture at least 24 hours in advance during drainage periods in irrigated fields. This methodology accounts for the influence of previously applied and planned irrigation volumes on moisture levels, drainage rates, and the critical thresholds required to prevent negative trunk shrinkage. Ultimately, this study demonstrates the viability of integrating efficient predictive models into agricultural IoT infrastructures, contributing to more sustainable water management in Valencian agriculture.
Source: Hispana
Author: Guerra Elena, Elsa
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
SDG: Zero hunger
Open Access status: green
OpenAlex ID: W7206488099
Cites:
Cited by:
Related to: