Journal of Biotechnology & Biomaterials
Open Access

Abstract

Most oilseeds (eg copra, palm kernels and groundnuts) need proper processing in mills before oil extraction to increase the yield of oil. The efficient and economical utilization of feed stocks is highly essential in oil producing industries. Coconut has the highest productivity and is less susceptible to abnormal climatic condition. The production of coconut oil and its by- products, raw and fried cake, is an important source of income for women in coastal areas of India. Hence, identification of optimal pretreatment conditions of coconut nut kernel is very important for high yield of coconut oil. The optimum processing conditions can be found by incorporating reliable and efficient statistical design methodologies such as central composite design (CCD), and ANN. Response Surface Methodology was used to conduct the experiments and experiments were designed according to CCD to study the effects of process variables such as Applied pressure, Pressing time, Roasting temperature, Roasting time and Moisture content. A simple, economical, and highly efficient model was developed to predict the yield of oil from coconut kernels in a hydraulic press. Artificial neural network (ANN) model was developed to predict the yield of oil from coconut kernels. The developed ANN was trained and tested with the experimental data obtained from CCD method. The results of ANN during training and testing were based on MSE. The results were compared with experimental data and it was found that the estimated oil yield from ANN model was able to predict the yield accurately with R value as 0.99.

Biography