<![CDATA[AbstractIn this paper a fuzzy optimization model of crude palm oil processing is presented. The model is focused on the minimization of oil losses and kernel losses during palm oil processing. The Mamdani fuzzy if then rules models are performed for each processing station in order to describe the relationships between the amount of oil losses and kernel losses with the processing variables which influence those losses. In conjunction with optimization method, genetic algorithm (GA) is applied to solve this problem. Using this method, the result shows that at 12 % of unripe fruits, the optimized sterilizing time is 110 minutes, steam pressure 2.9 kg/cm2, digesting time 21 minutes, digesting temperature 92.6 oC, screw press pressure 35 bar, CST and decanter temperature 90 oC, oil+NOS concentration in decanter feeding 55 %, rotor speed of ripple mill 1950 rpm, air velocity at LTDS I 10 m/s and at LTDS II 14 m/s. At these conditions, the minimum total loss is 1.522 % of the amount of fresh fruit bunch. This consists of oil loss is as 1.107 % and kernel loss is as 0.41 % of the amount of fresh fruit bunch.     Keywords: Crude Palm Oil; Qualitative Model, Optimization, Fuzzy If Then Rules, Genetic Algorithm]]>
Copyrights © 2005
