Research On Texture Detection System For Food Which Own Viscosity And Elasticity

The detection of food texture is an important way in the analysis of food organizational structure, which plays a significant role in food research and development, quality testing and processing facilities development. Equipment testing is a good substitute for sensory measurement, which is objective, efficient and easy opersation. However, there are common problems like empirical operation, single function, larger error in testing result compared with sensory testing in present testing equipment. This paper has developed texture testing system specialized for viscoelastic food materials with bionics techniques, computer technology and database technology used, which has overcome many deficiencies of present food testing.Firstly, in allusion to the requirement that texture from equipment testing ought to forecast texture from tast testing and on the basis that universal search from genetic algorithm could remedy the shortage of regression analysis, this paper has put forward the optimal texture prediction model theory based on genetic algorithm. Moreover, expression of correlation coefficient augmented matrix, converted expression of augmented matrix and expression of significant judgment between independent variables and dependent variables have been obtained with stepwise regression analysis in use. This paper has also defined all items that are needed in setting up the optimal texture prediction model, which provids the theoretic foundation for the realization of following systematic taste forecast.Secondly, bionic chew testing platform based on 6-UPS Parallel Mechanism has been set up according to human masticatory system structure. On the basis of sscannogram of average people’s chew trajectory, approximate function expression has been set up. In addition, the equation of motion planning expression of six connecting rods responsible for the motion og testing platform has been come up with. This paper, by dividing bionic chew trajectory into 8 parts and compiling MATLAB procedure, has successfully worked out precise expression of six connecting rods. This undoubtedly proves the efficiency of the bionic chewing movement trajectory planning theory and lays the foundation for the control function of systematic bionic chewing movement.Tirdly, this paper has also set up the structure and realized the function of texture testing system for viscoelastic food materials. The division of functional module, entire allocation, data constitution and organization modes of the whole system have been designed well according to the function and goal of the system. Visual C++ 6.0 as its platform, modes like main interace, sensory evaluation, taste prediction and motion control of testing platform of the system have beenestablished. As a result, many functions have been realized, for instance, the realization of trajectory planning expression, the setting of motion control parameter, data reading, the drawing of dynamic two-dimensional graph, the calculation and display of texture, the wizard type statistics of sensory testing data, the expression of optimal taste forecast, taste forecast as well as the read-write of data from database. Taking Access 2007 as the platform, data recording sheets have been designed, data of the system has been under perfect control.Finally, testing function and effectiveness of the system have been verified in the experience. In the experience, fudge, the representative of viscoelastic food was chosed as material. The three kinds of experiment parameter chosed are in different size, comprssion and testing speed which have a significant impact on texture testing. The result of the experience shows that in the tasting pattern of TPA with maximal force 25 N, optimal information of texture testing will be obtained with sample size being 10′10 ′10mm, compression being 40% and testing rate being 20mm/min. When 7 kinds of fudge are included in the experience under the same testing condition, analysis meter, system in this paper, sense organs all have been adopted for texture testing. The correlation analysis result, between sensory testing and another two methods, show that the correlation both system detection and sensory testing is significantly, and is more significant than the analyzer. The result verified the good function of system in this paper. Selected 7 kinds of soft candy and choose 3 groups in every kind of fudge, and used system and sensory do test to this sample. Then the system and SPSS software will use this data to build the taste prediction equation respectively. According to the comparison of two kinds of equation’ F value, T value and uncertainty factor, we all find that the equation of system build is better than the one of SPSS build. The result validate that the theoretical of construct optimal taste prediction equation is superiority. In addition, the system do test to another 3 kinds of soft candy, and calculate taste predictive value. According to the analysis of error between taste prediction value and sensory test value, we find that the average error less than 8%, standard deviation al within [1,2], this validate that the function of taste prediction equation is built by system is effective.