Penggunaan Analisis Regresi Linear Berganda untuk Mendapatkan Model Prediksi Respon Asam Lemak Bebas dan Dobi Hasil Rebusan Tandan Buah Segar Sawit

Abstract

Target of this research are : ( 1) Knowing physical result of steaming of Fresh Fruit Bunch; ( 2) Knowing the chemical of result of steaming of Fresh Fruit Bunch; ( 3) Finding model of prediction respon result of sterilization of Fresh Fruit Bunch ( 4 ) Determining time depth of steaming most precise to condition of raw material ( FFB ) to be sterilization. This research done at PT.Bio Nusantara Teknologi Bengkulu, by using horizontal type sterilizer. Research method the used  experiment which is  direct to be done at factory.  Multiple regression analysis to analyse data obtained, to the investigation of influence of free variable to variable of respon representing indicator efficacy of sterilization operation. Result of research indicate that treatment of variation : heavy classification of bunches,  fruit maturity and time depth steaming to have an effect on to respon result of sterilization perceived. Model of Prediction of met steaming respon as follows : Y1(Evaporation %) = 5,18 - 0,042 X1- 0,012 X2+0,082 X3 ; Y2 ( Fruit lose, g)=1629,96+8,14X1 - 3,22 X2 - 10,22 X3 ; Y3(Throughput apart mesokarp, g / men ) = - 0,748 - 0,297 X1 + 0,111 X2 + 0,336 X3 ; Y4 (percent heavy of mesokarp = 60,12 - 0,044 X1 + 0,018 X2 - 0,103 X3 ; Y5 ( Amount of kernel break = - 1,515 - 0,06 X1+ 0,035 X2 + 0,115 X3 ; Y6 ( FFA )= - 0,655 + 0,011 X1 + 0,029 X2 + 0,032 X3 and  Y7 ( DOBI) = 5,205 - 0,001 X1 + 0,01 X2 - 0,024 X3. Time depth steaming correct to braise small FFB, unrippe = 95 minute with score assess 7; Small FFB, mature = 95 minute with score assess 6; Small FFB over  mature = 90 minute with score assess 6. Time depth steaming correct for big FFB, unrippe = 100 minute with wight assess 6; Big FFB mature = 80 minute with score assess 5 and for big FFB, over mature = 95 minute with score assess 6.