Abstract
INTRODUCTION: 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. METHODS: 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. RESULTS: 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. DISCUSSION: 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.