Living Tree Moisture Content Detection Method Based on Intelligent UHF RFID Sensors and OS-PELM

Area of Science:

• Forestry Science
• Sensor Technology
• Machine Learning

Background:

• Accurate moisture content (MC) detection is crucial for living tree management and the advancement of forestry informatization.
• Current MC detection methods suffer from high energy consumption, low practicality, and poor sustainability.
• There is a need for non-destructive, efficient, and sustainable solutions for monitoring tree moisture levels.

Purpose of the Study:

• To design and implement an ultra-high-frequency radio frequency identification (UHF RFID) sensor system for non-destructive living tree moisture content (MC) detection.
• To develop a high-efficiency recognition model using deep learning for accurate MC prediction in forestry applications.
• To address the limitations of existing MC detection techniques in terms of energy consumption, practicality, and sustainability.

Main Methods:

• Implementation of a UHF RFID sensor system comprising passive tags for tree trunk mounting and a remote data processing terminal.
• Collection of data from living trees using a UHF reader.
• Development and training of an improved online sequential parallel extreme learning machine algorithm (OS-PELM) for MC prediction, featuring adaptive neuron network structures.

Main Results:

• The OS-PELM-based MC prediction model achieved a root-mean-square error (RMSE) of no more than 0.055 within a 1.2 m measurement range for living trees.
• The model demonstrated superior performance compared to other algorithms, with a mean absolute error (MAE) of 0.0225 and RMSE of 0.0254.
• The prediction model exhibited high precision, strong robustness, and broad applicability across the living tree dataset.

Conclusions:

• The designed UHF RFID sensor system effectively meets the demands of forestry Artificial Intelligence of Things (AIoT).
• The OS-PELM algorithm provides an accurate and robust method for non-destructive moisture content detection in living trees.
• This technology offers a sustainable and practical solution for improving forestry monitoring and management.