Unified moisture algorithm for improved RF dielectric grain moisture measurement

This paper presents a Unified Grain Moisture Algorithm based on measurements of the real part of the complex permittivity of grain at 149 MHz. The main goal of the method was to enable different moisture meter models to provide equivalent moisture predictions without calibration development. The database that was used to create and test the method included 6189 grain samples representing 53 grain types of US grain that were collected over a period of 6 years. The overall standard deviation of differences with respect to the air-oven method was 0.34% moisture. Methods for transferring grain moisture calibrations from the original 'Master' test cell to a smaller test cell were developed and tested. Three temperature correction functions with differing levels of complexity were evaluated. The algorithm is available as a public algorithm for commercialization by multiple manufacturers.