Design of fuzzy logic control system incorporating human expert knowledge for combine harvester

Many factors affect the yield loss in wheat harvesting with a grain combine harvester. Grain harvesting is a non-linear process, is of considerable complexity, and there is no mathematical model to describe the behavior of this complex system. In this paper, a fuzzy logic controller (FLC) incorporating human expert knowledge is designed for automatic adjustment and control of the harvester to achieve minimal grain losses especially at the position of straw walker and upper sieve. The FLC automatically adjusts cylinder speed, concave clearance, fan speed and forward speed of the combine based on the measured losses at straw walker and sieve sections. The designed FLC expert system consists two inputs and four outputs. Trapezoidal membership functions were selected for input fuzzy linguistic variables (straw walker and sieve losses), whereas fuzzy singletons were considered for the outputs. Based on human expert knowledge, six rules with logical AND operator and Mamdani implication are extracted. FLC was implemented in a programmable logic controller (PLC). Field experiments in two different irrigated or non-irrigated cultivated areas in order to evaluate the system. It was found the losses at the position of straw walker and upper sieve in the irrigated wheat cultivated area is much higher than the dry wheat cultivation area. Statistical analysis using t-test was also indicated a significant difference (p<1%) between loss mean in the combine equipped with the controller and the one without FLC.