多元统计分析技术在市售纯生啤酒一致性研究中的应用

Application of multivariate statistical analysis techniques in the consistency study of commercially available draft beer

中文关键词:  感官  风味  多元统计分析  主成分分析  聚类分析；纯生啤酒

英文关键词:sensory  flavor  multivariate statistical analysis  principal component analysis  cluster analysis  draft beer

中文摘要:

      目的 研究不同品牌和不同工厂市售纯生啤酒特征和产品质量一致性。方法 以国内3个品牌共计9个工厂1年内生产的不同批次纯生啤酒为研究对象, 采用气相色谱法测定啤酒的风味代谢产物并由国家级专业评酒委员进行样品感官评价分析, 以多元统计分析方法中的主成分分析(principal component analysis, PCA)、聚类分析(cluster analysis, CA)研究不同品牌纯生啤酒一年内生产的不同批次之间的产品特质和一致性。结果 研究结果显示3个品牌之间风味代谢物差异不大, 感官评价分析差异较大, B品牌一致性最差, A品牌次之, C品牌最好。B品牌出现日光臭比率为19.4%, 出现均值0.18以上纸板味的比率为48.4%, 缺陷量级为专业级国家评委可感知, 普通消费者难以察觉, 国内纯生啤酒风味评价良好。结论 多元统计分析方法可用于产品质量一致性评价, 不论是生产过程中的麦汁、发酵液等半成品还是啤酒成品控制中都能够发挥一定作用, 这些工具的运用可提升啤酒企业的产品质量一致性。

英文摘要:

      Objective To study the characteristics of different commercially available draft beer and the consistency of their quality in different batches produced by different breweries. Methods Beer samples from a total of 9 breweries which belonged to 3 brands produced within one year were selected as the research objects, the flavor metabolites in the beer were determined by gas chromatography, and the samples were evaluated and analyzed by the appraisal judge of beer. The multivariate statistical analysis methods, such as principal component analysis (PCA) and cluster analysis (CA) were used to investigate the product characteristics and consistency between different batches of draft beer produced by different brands within one year. Results The results showed that the differences in flavor metabolites between the 3 brands were not significant, but the differences in sensory evaluation analysis were significant. The consistency of Brand B was the worst, Brand A was the second, and Brand C was the best. Brand B had 19.4% of the light struck flavor, 48.4% of the cardboard flavor above the mean value of 0.18, which could be perceived by the appraisal judge of beer, but were difficult for ordinary consumers to detect. The flavor of domestic draft beer was good. Conclusion Multivariate statistical analysis methods can be used to evaluate product quality consistency, whether for control of products such as wort and fermented liquor during production or for control of finished beer. The application of these tools can improve the product quality consistency of beer enterprises.