Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
19556 | Food and Bioproducts Processing | 2007 | 8 Pages |
Back-propagation artificial neural network and response surface methodology were used to build a predictive model of the combined effects of molecular distillation's independent variables including evaporating temperature, feed flow rate and wiper rolling speed for the recovery of tocopherol from rapeseed oil deodorizer distillate. The optimum operating conditions obtained from the quadratic form of the response surface methodology and artificial neural network models were evaporating temperature 473 K, wiper rolling speed 150 r min−1, feed flow rate 90 ml h−1 when feed temperature 353 K and vacuum 0.02 torr. The results demonstrated a high predictive accuracy of artificial neural network compared to response surface methodology. The interior relationships between parameters are shown well by response surface methodology.