Microwave Measurement of Grain and Biomass
Abstract
Microwave measurement is a powerful method for mass prediction and grain detection. Existing systems are well adapted for small volume grain measurements, but cannot accommodate materials other than grain (e.g. biomass), which are larger in size or dimensional ratios. Therefore, an enlarged measuring system with unique design is needed. This new system’s large size improves measurement repeatability by capturing an average response for randomly aligned bulk samples. The system supports homogenous samples up to 0.5 m thick and 1 m square, and it has been validated from 1-50 GHz. By increasing the measurement system’s physical size relative to existing systems, it is now possible to measure the plants associated with these grains and seeds in a similar manner. Preliminary results for mixtures of grain and biomass are reported.
Having been validated for grain measurements, the One Meter Fixture is next used to collect phase shift and attenuation data for a variety of grain and oil seed samples (soybean, canola and corn). Using multiple variable linear regression analysis, a comprehensive clean grain mass estimation model was developed based on the dielectric properties of the grain samples derived from the S-Parameters at 13 GHz. Dielectric (ε') constant / properties and phase shift were introduced into the regression models and generated a grain mass estimation result with R2 values of 0.976, 0.977 and 0.989 for soybean, canola and corn samples, respectively. The results indicate that RF sensing technology has the potential to provide more accurate non-contact sensing methods for estimating grain mass in multiple precision agricultural applications.