Design of a Dynamic Grain Flow Model for a Combine Harvester

Nowadays, commercial grain flow sensors are available for each type of combine harvester. Based on this signal and other site-specific data, grain yield maps are produced, revealing local field characteristics. In contrast to the technology of these sensors, the data processing part is not optimized yet. Former studies show how significantly the combine harvester smooths feedrate variations and by this, bounds the accuracy of constructed grain yield maps. An analytical grain flow model is constructed based on physical insight and experimental data of selected machine sections. From this non-linear high order model, a low order linear model for the frequency band of interest is extracted. An optimal linear model which has algebraic compensation for the internal return loop, estimates the grain feedrate of the harvester. Once the dynamic behavior of the harvester is known in a simple, linear form, model-based grain yield estimations can be constructed, that may be used to improve the quality of constructed grain yield maps