| Abstrakt: |
Extracting groundnut oil from the groundnut seeds without modifying its quality become a critical task nowadays. This study utilized direct press method to extract groundnut oil from the seeds, then second-order response surface methodology (RSM) experiment is employed in conjunction with a five-level factorial Central Composite Design (CCD) for optimization. The interactions between the process factors are investigated, including pressure (A), groundnut size (B), Steam flow rate (C) and time (D). At a pressure of 80 MPa, a peanut size of 0.33 mm, a steam flow rate of 10 kg/h, and a time of 75 minutes, the maximum oil extraction efficiency of 55% is reached. Similarly, Saponification factor of 198 is reached at a pressure of 80 MPa, groundnut size of 0.3 mm, steam flow rate of 10 kg/h and a period of 60 minutes, whereas Iodine value 98 is achieved at a pressure of 80 MPa, groundnut size of 3 mm, steam flow rate of 10 kg/h and time of 75 minutes. The experimental R2 results show that the surface model prediction model is highly accurate, with an R2 value of 0.98. Overall, RSM in conjunction with the CCD will assist in identifying the critical operational parameters for extracting oil using a press type extraction equipment. The weighted K nearest neighbouring algorithm is also used in this work to predict the oil extraction efficiency (target output) based on the training data sets of pressure, groundnut size, Steam flow rate and time as input factors. [ABSTRACT FROM AUTHOR] |
