QUANTIFYING THE EFFECTS OF ADJUSTMENT ON THE FIELD PERFOJIMANCE OF A COMBINE CLEANER

Abstract

ABSTRACT Developing a technique for automatically adjusting a combine harvester has proven to be a very difficult problem. It appears to be one example from a class of process control problems in which there are wide variations in input conditions, many types of possible process adjustments, and little knowledge about how the process responds to changes in adjustments. The solution to the problem may lie in a system which integrates artificial intelligence techniques with traditional control theory, and is based upon both quantitative and qualitative data. A technique is proposed and demonstrated for gathering the needed type of quantitative information. Models were generated which describe how the combine cleaner responded to changes in fan speed, chaffer opening, sieve opening, and feedrate. Cleaner response was measured for five performance parameters: grain loss, tailings grain, tailings chaff, grain damage, and grain cleanliness. Additional information was gained about how the cleaner responded to the crop changes which occur within a daily harvest cycle. A technique is proposed for optimizing combine adjustments based upon an informed search strategy which utilizes quantitative models.