RFID-Based ZnO/TiO₂/SnS₂ Soil Oxygen Content Sensor Coupled with 1D-CNN-GRU Model: Classification for Predicting Soil...

Area of Science:

• Agricultural Engineering
• Materials Science
• Sensor Technology

Background:

• Soil oxygen is crucial for greenhouse crop yield and quality, impacting root respiration and nutrient absorption.
• Traditional soil oxygen monitoring methods suffer from limitations in cost, stability, accuracy, and anti-interference capabilities.
• Developing stable and precise monitoring systems is essential for effective greenhouse management.

Purpose of the Study:

• To develop a novel RFID sensor utilizing ZnO/TiO2/SnS2 nanocomposite materials for accurate soil oxygen monitoring.
• To integrate an intelligent algorithm (1D-CNN-GRU) to enhance sensor performance and data processing.
• To overcome the limitations of conventional monitoring techniques in greenhouse cultivation.

Main Methods:

• Fabrication of an RFID sensor on an FR-4 substrate using multilayer ZnO/TiO2/SnS2 nanocomposites.
• Implementation of environmental multifrequency encoding for multi-parameter perception and signal separation.
• Development and application of a 1D-CNN-GRU model for sensor data analysis and prediction.

Main Results:

• The sensor demonstrated improved response linearity (R^2 from 0.97 to 0.99) within the 5-25% oxygen range.
• Real-time dynamic monitoring accuracy increased by 3.16% compared to traditional methods.
• The classification task achieved 98.4% overall accuracy, with high precision (98.8%), recall (97.4%), and F1 score (97.7%).

Conclusions:

• The combined RFID sensor and 1D-CNN-GRU algorithm offer significant advantages for soil oxygen monitoring.
• This technology provides reliable data for soil health management and intelligent regulation of greenhouse crops.
• The developed system enhances crop growth optimization through precise environmental control.