切达奶酪挥发性特征风味物质的多典型相关融合分析

Dentification of Volatile Flavor Characteristics of Cheddar Cheese with Multi-CCA Fusion

切达奶酪在其成熟期内的感官品质级别与其挥发性风味组分之间存在相关性,而传统的特征风味组分相关性、贡献程度研究方法常依赖于仪器及人的主观因素制约,甚至无法摆脱化合物香气协同作用的束缚,存在很多缺陷,不能高质量快速的判断出特征风味物质对于不同奶酪级别的重要性以及贡献程度。针对这一科学问题,本文将切达奶酪挥发性风味物质组分与其感官品质级别的相关性展开探讨,对相关性分析的算法进行研究,提出一种多CCA(Canonical Correlation Analysis)融合的切达奶酪挥发性特征风味物质识别分析方法。经过实验得出,本方法在成熟期内淡味、中味、浓味切达奶酪中具有优势,优于CCA典型相关分析、PCA主成分分析以及PLS偏最小二乘法,能够更好的识别与切达奶酪淡味、中味、浓味3种感官风味级别相关性较强的风味组分。

There is a correlation between the sensory quality level of cheddar cheese and its volatile flavor components during the maturity period.Meanwhile,the traditional methods of correlation and contribution degree of flavor components often depend on the subjective factors of instruments and evaluators,and even unable to get rid of the bondage of compound aroma synergy,there are many defects,and it is impossible to judge the importance and contribution of characteristic flavorsto different cheese grades with high quality and speed.Base on this scientific issue,this paper discusses the correlation between the volatile flavor components of cheddar cheese and its sensory quality level,moreover studies the algorithm of correlation analysis.Proposing a multi-CCA(Canonical Correlation Analysis)fusion Cheddar Identification and analysis method of volatile characteristic flavor compounds in cheese.Through experiments,it is concluded that this method has advantages in mild,medium-flavor and mature-flavored cheddar cheeses during the maturation period,and is better than CCA canonical correlation analysis,PCA principal component analysis and PLS partial least squares method,and can better identify and cheddar.Meanwhile,the components of cheddar cheese which have strong correlation with mild,medium and mature flavor can be well classified.