Identification of the cleaning process on combine harvesters. Part I: A fuzzy model for prediction of the material other than grain (MOG) content in the grain bin

Operating a combine harvester can be described as very exhausting because of the high environmental temperatures, long working days and time pressures. In order to lighten the job load, automation technologies have been introduced. In this study, the focus is on the automation of the cleaning unit. A non-linear prediction model for the material other than grain (MOG) content in the grain bin has been established by means of fuzzy modelling techniques. Different model structures were evaluated and best results were obtained when the input data space consists of two input variables (fan speed and load on the upper sieve section) and is subdivided into three local linear sub-models. It was shown that the cleaning section settings such as lower and upper sieve openings had a minor effect on the content of MOG in the grain bin when compared to other variables such as fan speed and the loadings by chaff, straw and grain on the upper sieve. High MOG values can be expected when both the loadings on the upper sieve are high and the fan speed is low. High fan speeds will result in lower MOG values, independent of the load on the upper sieve, but they can also give high losses.