# Non-destructive detection of Phyllostachys edulis under salt stress using UHF RFID based on Cole-Cole model optimization algorithm

## TL;DR

This paper introduces a non-destructive method to detect salt stress in Phyllostachys edulis using UHF RFID and an optimized Cole-Cole model for improved accuracy.

## Contribution

A novel C-T-PSO algorithm is developed to optimize the Cole-Cole model for salt stress detection in bamboo.

## Key findings

- The C-T-PSO-Cole-Cole model achieved over 93% accuracy in salt stress detection.
- The model outperformed six other swarm intelligence algorithms in performance metrics.
- Non-destructive diagnosis of salt stress in Phyllostachys edulis reached 95.3% accuracy.

## Abstract

Salt stress disrupts cellular osmotic balance in Phyllostachys edulis, alters leaf ion distribution and thereby affects dielectric properties. To meet the demand for non-destructive salt stress detection, this study proposes a diagnostic method integrating multi-physics field coupling characteristics.

Based on the mechanism of salt stress regulating ion concentration in cell sap, a Cole-Cole dielectric model detection framework was constructed by analyzing intrinsic correlations between RFID backscattering signal features and medium dielectric properties. An improved Particle Swarm Optimization (C-T-PSO) algorithm employing Chebyshev chaotic mapping for population initialization and t-distribution dynamic perturbation mechanism was developed to synergistically optimize Cole-Cole model parameters.

Experimental verification showed the C-T-PSO-Cole-Cole hybrid model exceeded 93% in all core metrics (accuracy, precision, recall, F1-score). Comparative experiments with six swarm intelligence optimization algorithms confirmed the model's comprehensive superiority. Convergence curve analysis based on standard test functions demonstrated faster and more stable convergence of the C-T-PSO algorithm. The final model achieved non-destructive diagnosis of salt stress in P. edulis using UHF RFID technology with 95.3% accuracy.

The hybrid model provides an effective real-time monitoring tool for salinized soil management in bamboo forests, validating the feasibility of salt stress detection through dielectric property analysis.