Apilikasi Matematika Linear Dalam Penyusunan Formulasi Optimum Pencampuran Kopi Arabika Gayo (Coffeea Arabica. L)

Abstract

Coffee is an Indonesian commodity that has a high selling price. Gayo Highland is known for arabica coffee plantation site in Aceh, where commonly three local cultivars are cultivated. These are Timtim or Gayo 1, Borbor or Gayo 2 and Ateng Super. All three cultivars complied as specialty coffee with cuptest score 84,50 (Timtim), 85.25 (Borbor) and 85.50 (Ateng Super) and each cultivar has distinctive aroma and flavour. Blending is one common practice before coffee brewing in order to optimize the cupping quality of coffee. This study aims to optimize the total cup test of blended coffee from three local cultivars of arabica coffee over 85 by using simple linear mathematics models. The hypotheses was a predicted cuptest score is equivalent with the laboratory cuptest score. Based on previous research, seven formulations were set up by using three local cultivars. Then two linear models were created based on specific constraints such as attribute score and cultivar ratio. The seven formulations are calculated in two assigned linear models as two replications. First model produced predicted score in between 84,15-85,38 and mean 84,88 whilst second model has a score range 84,07-85,13 with mean 84,80. From all seven formulation, in these two linear models R3 and R7 has predicted cuptest score over 85. By using t-test, there is not significant difference available within the laboratory score and predicted score. It can be said the hypotheses is accepted and these two linear models could be used separately and or intentionally to predict cuptest score of blended coffee before blending is performed.