Modeling And Application Of The Viscosity Curve Of Wheat Flour

With the deepening of the wheat quality research, people gradually realized the importance of the wheat starch in food processing and quality breeding process. Lots of studies indicated that wheat starch has a big effect on wheat processing quality, edible quality and starch products quality. Most research results showed that wheat starch gelatinization characteristics were the important index of the wheat starch quality, under the influence of genotype, environment and their interactions. Hence, improving wheat starch gelatinization characteristics has become an important target of wheat quality breeding. However that many research results of wheat starch gelatinization characteristics are contradictory indicates it remains to be further studied according to the analysis of existing RVA characteristic parameters. In order to further expand the appliance of RVA characteristic parameters in the breeding process, we do quantitative analysis of the gelatinization characteristics of wheat varieties in this paper.Data analysis application in the field of scientific research has occupied a pivotal position and most research results depend on the data analysis correctly, among which data fitting has become increasingly popular. In various scientific research experiment involving a large amount of data, finding out the laws between the data, establishing proper mathematical model, carrying on the quantitative description, are the main tasks of the data fitting. In this paper, we summarized some commonly used methods of data fitting, especially for nonlinear data fitting, analyzing their data processing and their characteristics. The basic idea of nonlinear regression is to require the sum of the squares of the difference between each point on the curve fitting and the original data minimum, selecting the proper fitting method according to the characteristics of data.RVA changing over time in the heating process is characterized by a typical curve relationship which can be described with a function, thus can explain the change process of RVA more specific. Through a large number of trial and screening, we found an equation with a high fitting degree. Based on the improved contraction-expansion algorithm, this equation can get the optimal parameter estimation starting from any initial values. In this paper we do the RVA curve fitting of63wheat materials with a fitting degree of99.5%, viscosity differences performing on the parameters of the equation.There are many articles on RVA characteristics research, most of which on the influencing factors of starch gelatinization characteristics, few on quantification analysis of the pasting properties. Moreover, it is the existing secondary data of RVA parameters (setback, breakdown) that makes further analysis of them(the original values) difficult and that meantime causes structure matrix discontent or pathological. Using RVA function relationships changing over time, you can get some new information points (new characteristic values) which is not the linear combination of the original data and will not cause above problems, thus make the statistical analysis going-on on the basis of the matrix array. These new characteristic values amplify and enrich the information of the starch viscosity, and maybe contain crucial information on flour quality. We have to take the first and second order derivative of the fitting equation to calculate the viscosity and the time of the RVA curve at turning points. The new RVA characteristic values proposed in this paper is based on the old ones, combined with the calculated viscosity and the time at the turning points, thus ensuring the integrity of starch viscosity information. For further research on starch viscosity properties, as well as mutual relationships between other major quality traits of wheat, this paper provides a theoretical basis.Clustering analysis is a kind of method of studying individuals based on its characteristics with the purpose of grouping similar things. Its principle is that the individuals in the same class have large similarity, while the ones in the different classes have large differences. In this paper we choose the minimum sum of squares within one group (MinSSw method) as the clustering method, this method can effectively adjust the initial grouping of individuals to a reasonable classification, and has good robustness. In order to further modified K-means, MinSSw considers the minimum sum of squares in one group that means the maximum sum of squares between groups as the reasonable classification standard, the clustering steps similar to K-means. In this paper, according to the original and new characteristic values in the standardization respectively, the63samples are divided into five types. By calculating their respective lambda statistics, we can achieve a better clustering effect based on the MinSSw method. In the meantime, by comparing two graphic clustering results, the new characteristic values can get better clustering effect than the original ones.Principal component is a linear combination of the original variables, it is an improvement on the original variable information which does not increase the total amount of information, also not reduce them. The cumulative percentage in the total variance (the cumulative contribution rate) decides how many principal components should be kept which indicated how much information they contained. This article selected the first three principal components with the cumulative variance contribution rate of99%, illustrating the three ones have fully reflect the quality traits of wheat, so9RVA characteristic parameters can be compressed to3principal components. Then clustering analysis of the three principal components, the test materials in the same group are very close in three dimensional space which helps the visual display of the result. In addition, the clustering results of the new parameters through principal analysis are consistent with those not, which indicated the three selected principal components could totally summarize the information of starch viscosity.