CCSR Sensor Design Project for Moisture Monitoring
in Grains
Microwave resonator sensor, relative humidity, grain size, CCSR, dielectric characterization, multivariate regression.
This work presents the development of two planar microwave resonator sensors based on Complementary Circular Spiral Resonators (CCSR) for estimating the moisture content of bean grains. Two sensor configurations were proposed, one with a single resonator and the other with two resonators, both implemented on a U-shaped microstrip transmission line. The sensors were designed and optimized through electromagnetic simulations using the High-Frequency Structure Simulator (HFSS) software, manufactured, and experimentally validated using samples subjected to different moisture levels. The results demonstrated that increasing moisture content causes a shift in the resonance frequency, a reduction in the magnitude of the S21 parameter, and a decrease in the quality factor. However, it was found that these parameters, when analyzed individually, do not adequately discriminate all the investigated moisture levels. To overcome this limitation, multiple spectral descriptors were extracted, and multivariate regression techniques were applied. The results showed that the multiparametric approach provides better predictive capability than the use of individual parameters, demonstrating the potential of the developed sensors for non-destructive monitoring of grain moisture.