Ph.D. Dissertation:

Evaluation and modeling of conjugated linoleic acid (CLA) changes in production process of dairy products using fuzzy inference system

The present study was designed to investigate the seasonal changes of fatty acid compounds, especially CLA in milk and its products using the gas chromatography approach. In addition, the effect of various processes on such compounds was studied. Using the Mamdani-type fuzzy inference under the if-then principle, the estimated CLA induced by the implemented conditions on time and temperature was simulated in the fuzzy logic model of MATLAB software.Moreover, the milk fat’s melting DSC profile was determined in four seasons in order to investigate the effect of fatty acid’s compounds on melting features. The results indicated that in among milk and dairy products, the CLAc9-t11 isomer was lowest (0.01) in spring’s intact milk and summer’s butter, and highest (0.08) in winter’s sterilized milk. As for the CLAc10, t12 isomer, the lowest amount (0.38) was detected in summer’s sterilized milk and the highest amount (0.59) was found in summer’s intact milk and spring and summer’s sterilized milk. The total CLA was lowest (0.42) in winter’s sterilized milk and highest (0.65) in spring’s sterilized milk. Seasonal changes showed no effect on the amount of isomers and the total amount of CLA in milk and its products within a 95% confidence interval, nor was there any significant difference among the four seasons in this respect. Regarding changes in the total CLA value from intact milk to the products, the pattern was descending in sterilized milk and yoghurt, ascending in butter, and unchanged in cheese. On the other hand, differences observed in the total CLA amount and its isomers were insignificant as far as the different production processes of the products under investigation are concerned. Through the measurement of milk fat’s CLA during heating, it was shown that CLA was prone to a greater degree of decomposition and oxidization following a temperature increase from 70?C to 130?C. In the kinetic analysis of CLA’s thermal degradation, the results showed that CLA thermal degradation follows the zero equation model. With respect to the description of CLA degradation’s temperature dependence in milk fat, the Arrhenius equation yielded an activation energy estimation of 23.10 KJ/M. The predicted CLA amounts based on the fuzzy logic model as opposed to empirical amounts induced a correlation coefficient exceeding 0.99, which proves the accuracy involved in the estimations within the fuzzy logic model. The melting curves of differential scanning calorimetry showed three conspicuous melting peaks owing to different triglycerides with different amounts of fatty acids in season-specific samples. Melting temperature in low and high melting fractions was highest in winter and lowest in summer. Melting temperature in medium melting fraction was highest in fall and lowest in summer.

Keywords: CLA, Fuzzy logic, kinetic, Differential scanning calorimetry, Milk and dairy products