Test Research Of Beer Based On Electronic Nose/Tongue

Beer has been one of the most popular products among consumers for its good taste and rich nutrition. On the other hand, some of the status such as appreciation level and quality requirement are increasing. At present, testing and evaluation methods of beer quality include traditional physical and chemical index detection method and the sensory evaluation method. The former is mainly used to review of some ingredients in beer; the latter is to evaluate the specific sensory quality of beer. Both of the two detection methods this method were have some defects, such as the evaluation time is long while the efficiency is low and so on. Therefore, a rapid detection method for beer quality is of great significance.This paper designed and built a fusion system of electronic nose and electronic tongue which was used for the detection of beer quality. The electronic nose system was mainly designed and formed from the four parts of sensor array, gas path, data acquisition and soft part. The electronic tongue system was adopted of a SA-402 B electronic tongue which used of five basic taste sensors. The feature level fusion method of information fusion was used to fusion of electronic nose system and electronic tongue system.The fusion system of electronic nose and electronic tongue was applied to test of five different brands of beer. The principal component analysis, hierarchical clustering and support vector machine methods were used to analyze the characteristic data which were collected by electronic nose system, electronic tongue system and the fusion system of electronic nose and electronic tongue, to observe the differences between different brands of beer and to compare the recognition effect of different systems. The results showed that all of the electronic nose system, electronic tongue system and the fusion system could realize the quickly recognition of beer brands, and the recognition result of the fusion system was the best.For the complete representation of fuzziness and randomness of beer sensory evaluation because of the sample differences, physiological and physiological factors, the fusion technology of electronic nose and electronic tongue was used for automatic evaluation of beer aroma, taste and the overall flavor sensory information. A mathematical model which freely switched between the concept of qualitative and quantitative values was built. The fuzzy neural network was trained with the characteristic information collected by the sensor coupling array as input, with the characteristic value being generated from the conversion of the cloud model of sensory evaluation as output. The error rate of the aroma and taste fuzzy information evaluation is between 0.41% and 9.93%. The effect of this system is good, as the error rate of the overall flavor fuzzy information evaluation is between 0.48% and 3.94%.