The Design And Implementation For A Comprehensive Analysi’s System Based On The Wine-based Data

With the improved quality of life, wine became very popular drink, wine quality measurement, analysis, evaluation for wine making process improvements, and improving the quality of life has a very important significance. In order to further analysis of wine, wine products we delve into the underlying data, based on the design and implementation of a comprehensive data base wine product analysis system. The main functions of the system are as follows.Features one is the difference detection. The quality of the wine tasters tasting scoring mainly through given. As sommelier scoring preferences are different, given the score differences. This system is the single factor analysis of variance, respectively, of two tasters scores for the group and the differences between groups for significance analysis. Data show significant differences within the group between the two groups was not significant red wine, white wine is remarkable. To evaluate the results of the evaluation of each group’s credibility, we have established an evaluation model based on analysis of variance, the data show a smaller variance of the second group, the second group of the evaluation results more credible.Features two is that wine grapes are graded. In order to brew good wine, selected high-quality grapes is essential. We first carried out using the confidence interval method wine appraisal heterogeneity analysis through corresponding algorithm fixes wine appraisal scoring, making scoring difference was no longer significant. Also assume not consider wine production process, storage, transportation and other effects on the quality of the wine, with the corrected score values will correspond grapes into class, namely excellent, in general, inferior. Secondly, the use of non-dimensional normalization method, the physical and chemical indicators of grape for data processing, and then using a principal component analysis method, obtained19principal components, the cumulative contribution rate of more than86%. Through each of the main components of the principal component analysis, the grape is the most representative indicators of alcohols, esters, acids, aldehydes and enzymes. Finally, multiple regression analysis, given the quality of the principal component regression equation and grapes in order to establish the physical and chemical indicators of the use of grapes for wine grape quality relationship, the test sample tests show that the model results are correct.Features three is to find grape and wine links between physical and chemical indicators. We first correlation analysis method to wine grapes and wine physicochemical indicators correlation analysis and comparison, generates correlation coefficient surface chart, select with primary or significantly associated with wine grapes and wine physical and chemical indicators, combined with information on the literature to determine its relevance causes. Secondly, the grape and wine composition were compared, gives the grapes ingredient in the brewing process of migration and transformation result. By analyzing the results:grape anthocyanins of red wine is the key color substance found in the skins of grapes; wine grapes in phenols and ketones generated by chemical reactions and chemical substances in wine affect the physical properties of the wine, Grape esters, alcohols remain stable in the brewing process, the taste and aroma of the wine source of dry matter and soluble solids content of wine mainly affect the brightness, color and other physical indicators.Features four is analyzing the quality of wine with grapes and wines physicochemical properties linked. We begin by physical and chemical indicators of the non-dimensional treatment to obtain a normalized data matrix, and then through all the physical and chemical indicators of wine and grapes principal component analysis, each drawn19principal components. Thereby establishing a regression wine quality evaluation model, the final prediction using experimental results of the evaluation sample was tested to verify the correctness of the model.Finally, according to the above model, we use C#as the implementation language to visual studio2012as a platform to achieve a wine analysis system. The system is mainly to achieve our above functions, the interface includes five aspects, namely, documents, functions, basic materials, tools, Window, and Help. Software design and implementation, we can load our model to the real system, it has a better application prospect.