【摘要】 目的：利用低场核磁共振技术结合指纹图谱，判别分析方法转换数据，将蜂蜜样品与标准数据库进行对照匹配，获得指纹鉴别区域的差异信息，从而能够快速鉴别蜂蜜的种类。方法 ：采用低场核磁共振仪对麦卢卡蜂蜜、枇杷蜂蜜、油菜蜂蜜、枣花蜂蜜施加指纹谱脉冲序列，获得了具有指纹特征的磁共振指纹谱，运用Matlab将图谱转化成42个数据变量，在SPSS平台建立Fisher判别分析模型。结果：通过对Matlab转换的蜂蜜指纹图谱数据建立判别模型，按照麦卢卡蜂蜜组、枇杷蜂蜜组、油菜蜂蜜组和枣花蜂蜜组分析，总的正确率达到91.1%。结论：结果表明，可以通过低场核磁共振指纹谱技术对上述4种蜂蜜种类进行有效鉴别，为蜂蜜的快速检测提供了新的思路，证明了核磁共振弛豫指纹谱方法对蜂蜜的鉴别检测的巨大潜力。更多还原

【Abstract】 By using low-field nuclear magnetic resonance technology combined with fingerprint spectra and discriminant analysis methods to convert data, the honey samples were compared and matched with a standard database to obtain differential information in the fingerprint identification region, thereby enabling rapid identification of honey types. Low-field nuclear magnetic resonance spectrometers were used to apply pulse sequences for fingerprint spectra on Manuka honey, loquat honey, rapeseed honey, and jujube flower honey.Magnetic resonance fingerprints with characteristic features were obtained and transformed into 42 data variables using Matlab. A Fisher discriminant analysis model was established on the SPSS platform. By establishing a discriminant model based on the Matlab-converted fingerprint spectral data of honeys from Manuka group, loquat group, rapeseed group and jujube group respectively analyzed; the overall accuracy rate reached 91.1%. The results demonstrate that low-field nuclear magnetic resonance fingerprint spectroscopy can effectively identify these four types of honeys and provide new insights for rapid detection of honeys.更多还原