A genetic input selection methodology for identification of the cleaning process on a combine harvester, Part II: Selection of relevant input variables for identification of material other than grain (MOG) content in the grain bin

Abstract

The cleaning process on a combine harvester is a complex process that is influenced by a wide range of parameters such as machine settings, field and crop-related parameters, etc. Because of the high time pressures combine drivers have to deal with, optimal settings for the cleaning section are usually only estimated once for each crop. As a consequence, differences in temporal and site-specific conditions are neglected. No recent literature is available that considers the interaction between the settings of the cleaning section (like e.g. fan speed, lower sieve opening and upper sieve opening) and the material other than grain (MOG) content in the grain bin, which is, however, an important performance parameter of the cleaning shoe. In this study, a combine harvester was equipped with extra sensors that could contain valuable information necessary to predict the performance of the cleaning section. A non-linear genetic polynomial regression technique was used to rank the pool of potential sensors as possible regression variables for a prediction model of the MOG content in the grain bin. This model is important for the automation of the cleaning shoe. Results showed that the MOG content in the grain bin is influenced non-linearly by differences in the amount of biomass on the sieve section and the fan speed, which are also correlated with each other.