Study On Detection Of Difference In Fenjiu Quality And Automatic Blending Technology

As a typical representative of fragrant liquors,Fenjiu is unique in its pure aroma and clean aftertaste,boasting more than 1,500 years of glorious history and rich cultural connotations.Although the market share of Fenjiu has risen in recent years,the taste of the flavour has remained unstable.The most important reason is the lack of technological innovation.Therefore,the topic takes Fenjiu as the research object.The main research content is as follows:The correlation between sensory evaluation,GC-MS detection technology and z Nose ^TMM electronic nose test results was discussed.The correlation between electronic nose and GC-MS was more than 95%.The effective characteristics of electronic nose was extracted to establish a prediction model of sensory scores.The coefficient of determination reached more than 90%.Through the use of the zNose^TMM electronic nose for the detection of different types of Fenjiu,the acquired data combined with the corresponding analytical methods can achieve differentiation of different types of Fenjiu.At the same time,it is found that the variable that contributes significant difference to the quality difference is consistent with the characteristic variable that contributes a lot to the aroma of Fenjiu.It is proved that zNose^TMelectronic nose is reliable and accurate as the testing instrument of Fenjiu.For the three kinds of finished Fenjiu A,B,and C shipped in different months of the same year,z Nose^TMM electronic nose and GC-MS detection technology were used to detect Fenjiu in each finished product.The sensory score regression model was used to predict the scores.The combination of electronic nose detection data and principal component analysis found that the integrated discrete proportions of finished Fenjiu A,B,and C all reached over40%.GC-MS detected the seven major aroma components in finished Fenjiu,maximum aroma content fluctuations reached 12.82%.Using the electronic nose detection data combined with the sensory score prediction model,the average variance and the variance were compared.The maximum variance of the sensory score was 69.5.The analysis of the three types of quality evaluation methods for finished Fenjiu A,B,and C shows that there are differences in the quality of the finished Fenjiu in different months of the same category.Compared with the products with the least difference in quality,finished Fenjiu B,the product with the greatest difference in quality is the finished Fenjiu A.The zNose^TMM electronic nose was used to detect the blended raw wines of finished Fenjiu B,and combined with corresponding mathematical statistics methods to analyze the test data.Different batches of raw wine were found to have different degrees of difference.Some of them were automatically grouped into three categories,and some were automatically grouped into two categories.Through the analysis of the mean,standard deviation,and coefficient of variation of the electronic nose detection variables,the characteristic variables that greatly contribute to the quality differences of raw wine were extracted.In order to solve the problem of the inconsistency in the taste of the product,the LABVIEW was used as a control platform,the zNose ^TMM electronic nose was used as an online detection method,an automatic alcohol blending system based on artificial neural network blending algorithm.Based on the design of this system,hardware assembly and software program writing and debugging are completed.By using the blended raw wines with different variability,an automatic blending experiment was conducted using a stable finished Fenjiu B as the target wine.Using the electronic nose,GC-MS detection technology and sensory score prediction model respectively,the three Fenjiu quality evaluation methods confirmed the consistency of the blended wine and the target wine,indicating that the design and construction of the automatic alcohol blending system can achieve the desired goal better.It proved that the system is reliable.