Establish Yoghurt Texture Parameter Prediction Model For Sensory Properties

For establishing a rapid method of set yogurt to complement the quality evaluation system, both texture test and sensory evaluation were selected and optimized. The two common methods were combined through the prediction models of sensory indictors built by partial least squares regression (PLS).In the texture optimization, deformation, test speed, interval and probe were selected to estimate the impact of set yogurt texture profile. Based on the influence, sensitivity of the Texture Analyzer and the capacity of discriminate different samples, the optimized test conditions were determined. The results showed that hardness and adhesiveness were proportional to deformation; adhesiveness was inversely proportional to test speed and other parameters increased differently as the increasing test speed; extension of interval time enlarged cohesiveness rather than others; the values of all texture parameters increased with the increasing diameters of the probe.20%deformation,60mm/min test speed without interval was determined as the optimal test condition. It was found that12.7mm probe was the most sensitive to distinguish concentration discrepancy of set yogurt and the least discrepancy was0.8%.In order to reflect consumers liking to yogurt products visually, sensory evaluation was also chose and optimized. The paper combined the weighting method and the clustering analysis method to evaluate the weight and correlation of surface water, spoon impression, sour, sweet, firmness, smoothness, adhesiveness, thickness (mouth) and palatability which were contributed to set yogurt sensory quality. The practical application values of preliminary selected sensory attributes were estimated through the questionnaire. Questionnaire investigation indicated that all consumers in different ages, sex, and incomes were fond of smoothness, adhesiveness, surface water, firmness and palatability differently, which were separate categories and the common weights were0.061,0.137,0.120,0.169,0.257. They could be selected as sensory indicators to evaluate set yogurt. For combining texture test with sensory evaluation availably in appraisal of yogurt quality, based on30data groups resulted from texture and sensory test of set yogurt made from various milk concentrations, culture style and fermentation time, five prediction equations for the above selected sensory indicators were established by PLS and of statistical significance (the fit degree was greater than0.85, p<0.05). The models can be useful tools for estimating yogurt quality rapidly and accurately and complement the yogurt quality evaluation system in some degree as soon as they were checked by some novel samples obtained from altering sugar, culture types and the homogeneous pressure, which was no significant difference between sensory actual and predicted values.