Oroian Mircea, Ropciuc Sorina, Paduret Sergiu, Sanduleac Elena Todosi
Faculty of Food Engineering, Stefan cel Mare University of Suceava, Suceava, Romania.
J Food Sci Technol. 2017 Dec;54(13):4240-4250. doi: 10.1007/s13197-017-2893-0. Epub 2017 Oct 9.
The purpose of this study was to investigate the physico-chemical properties (free acidity, pH, a, ash content, moisture content, color (L*, a*, b*, hue-angle, chroma and yellow index), fructose, glucose and sucrose content) and textural parameters (viscosity, hardness, adhesion, springiness, cohesiveness, chewiness and gumminess) of 50 samples of honey of different botanical origin (acacia, polyfloral, honeydew, sunflower and tilia). In order to achieve the authentication of the honey samples analyzed, their data have been subjected to linear discriminant analysis (LDA) and principal component analysis (PCA).The PCA and LDA have proved the possibility of honey authentication using the physico-chemical and textural properties. LDA classified correctly 92.0% of the honeys based on their botanical origin, using the cross validation. In the LDA projection, the textural parameters (chewiness, hardness, cohesiveness, springiness) dominated the two functions.
本研究的目的是调查50个不同植物来源(刺槐、多花、甘露、向日葵和椴树)蜂蜜样品的物理化学性质(游离酸度、pH值、a值、灰分含量、水分含量、颜色(L*、a*、b*、色相角、色度和黄度指数)、果糖、葡萄糖和蔗糖含量)以及质地参数(粘度、硬度、粘附性、弹性、内聚性、咀嚼性和胶粘性)。为了对所分析的蜂蜜样品进行鉴定,已对其数据进行线性判别分析(LDA)和主成分分析(PCA)。PCA和LDA已证明利用物理化学性质和质地特性进行蜂蜜鉴定的可能性。使用交叉验证,LDA根据植物来源正确分类了92.0%的蜂蜜。在LDA投影中,质地参数(咀嚼性、硬度、内聚性、弹性)在两个函数中占主导地位。