Analysis of Stability and Type Independence of Three Density-Independent Calibration Functions for Microwave Moisture Sensing in Shelled and Unshelled Peanuts

Abstract

A microwave dielectric method was used for nondestructive and rapid determination of moisture content in shelled and unshelled peanuts of various types from transmission measurements of their relative complex permittivities in free space at 23C between 5 and 15 GHz. Moisture content was estimated, independent of bulk density, with three density-independent calibration functions and compared to standard oven moisture determinations; two of these functions are permittivity-based, and the other is expressed in terms of attenuation and phase shift. The effectiveness and stability of these three functions for type independence were evaluated over broad ranges of frequency, moisture content, and bulk density. While the performance of each function with individual type calibrations was reaffirmed, statistical analysis also showed high coefficients of determination in predictions with the combined type-independent calibrations. Therefore, with microwave moisture sensing technology, calibration equations can be used to accurately predict moisture content in peanuts with insensitivity to type; which is a characteristic lacking in today's commercial moisture meters.