Effect of temperature and starch concentration on the creep/recovery behaviour of the grape molasses: modelling with ANN, ANFIS and response surface methodology


Toker O. S. , DOĞAN M.

EUROPEAN FOOD RESEARCH AND TECHNOLOGY, cilt.236, ss.1049-1061, 2013 (SCI İndekslerine Giren Dergi) identifier identifier

  • Cilt numarası: 236 Konu: 6
  • Basım Tarihi: 2013
  • Doi Numarası: 10.1007/s00217-013-1959-0
  • Dergi Adı: EUROPEAN FOOD RESEARCH AND TECHNOLOGY
  • Sayfa Sayıları: ss.1049-1061

Özet

In this study, the effect of starch concentration (5, 7.5 and 10 %) and temperature (60, 70 and 80 A degrees C) on the creep and recovery behaviour of grape molasses was investigated. Adaptive neuro-fuzzy inference system (ANFIS) and artificial neural network (ANN) models were established for the prediction of the compliance values (J(t)) based on temperature, starch concentration and time of the creep or recovery phases. The root mean square error, mean absolute error and R (2) values were used for the comparison of the models which showed that the ANFIS model performed better than the ANN model for the desired purpose. The Burger model fitted the J(t) versus time data with R (2) values ranging from 0.987 to 0.999. Response surface methodology (RSM) was performed to investigate the dependency of the creep (G (0) , G (1) , n (0) and n (1) ) and recovery (J (KV) , B, C, J (max) , J (a) and % recovery) parameters to temperature and starch concentration. As a result of this study, it was observed that deformation of the grape molasses samples increased with decrease in starch concentration and increase in temperature. The gel strength (S) values of the samples were also calculated and modelled by RSM. As increase in starch concentration caused an increase in S value, there was an inverse proportion between the temperature and S value.

In this study, the effect of starch concentration (5%, 7.5% and 10%) and temperature (60 °C, 70 °C and 80 °C) on the creep and recovery behavior of grape molasses was investigated. Adaptive neuro fuzzy inference system (ANFIS) and artificial neural network (ANN) models were established for the prediction of the compliance values (J(t)) based on temperature, starch concentration and time of the creep or recovery phases. The root mean square error (RMSE), mean absolute error and R2 values were used for the comparison of the models which showed that the ANFIS model performed better than the ANN model for the desired purpose. The Burger model fitted the J(t) versus time data fitted to the with R2 values ranging from 0.987 to 0.999. Response surface methodology (RSM) was performed to investigate the dependency of the creep (G0, G1, n0 and n1) and recovery (JKV, B, C, Jmax, J? and % recovery) parameters to temperature and starch concentration. As a result of this study, it was observed that deformation of the grape molasses samples increased with decrease in starch concentration and increase in temperature. The gel strength (S) values of the samples were also calculated and modelled by RSM. As increase in starch concentration caused an increase in S value, there was an inverse proportion between the temperature and S value.