【作者】 廖步岩；

【导师】 张正竹；

【作者基本信息】 安徽农业大学 ， 农产品加工与贮藏工程， 2009， 硕士

【摘要】 近红外光谱(NIRS)是介于可见光谱区和中红外光谱区之间的电磁波。近年来,近红外光谱技术以其在相对准确分析的基础上,兼有快速、简便等优点而迅速发展成为一种新兴的分析与研究手段。茶叶品质属性的快速准确甄别是当前茶叶行业亟待解决的一个重要课题。现有的茶叶品质属性甄别方法主要有感官审评和指纹认证技术(包括生物指纹图谱和化学指纹图谱)两种评价方法,这两种方法都不同程度地存在局限性。近红外光谱技术是一种新兴的分析与研究手段,通过提取茶叶全光谱信息,借助数学方法建立能充分反映全谱区光谱特征与茶叶品质专有属性关系的数学模型,找出被众多错综复杂的共性所掩盖着的专属特征,就有可能定性或定量地描述不同样品谱图间的相似程度,从而对样品快速而客观地分型划类,这是NIRS技术可以用于评判茶叶品质专有属性甄别的原因所在。本文选用以黄山毛峰为主的66种茶叶样品,探讨了利用近红外光谱分析技术,结合不同的化学计量学方法(偏最小二乘法、标准法和因子法),对黄山毛峰茶进行快速无损定性检测,进而判定类别的新方法。结果表明,基于提取的茶叶近红外光谱数据,利用标准法结合因子法综合定性效果优于偏最小二乘法(PLS),可以直观地将66个来自不同产地、品种和加工工艺的茶叶样品进行准确的分类判定,特别是可以准确地定性区分黄山毛峰的不同产地。同时论文对66种茶样中的11种主要组分进行了检测(国家标准检测方法),再结合近红外光谱构建了定量分析模型。在建模过程中,光谱预处理方法采用消除常数偏移量、一介倒数、二阶导数、矢量归一化、减去一条直线、最小-最大归一化、多元散射校正(MSC)、一介倒数+减去一条直线、一介倒数+矢量归一化、一介倒数+多元散射校正(MSC)或没有光谱预处理等11种光谱预处理方法,随机选取70%的样品建模,剩余30%样品用于对模型进行外部检测。结果显示水分、水浸出物、粗纤维、咖啡碱、ECG、EGCG及儿茶素总量等7种指标的相关系数R2达到90以上,分别为94.73、93.52、95.37、97.39、96.73、95.24和98.21,C和EC的相关系数R2达到84.89和86.34,达到了定量分析茶叶组分的要求;而总灰分和EGC的相关系数R2只有71.61和56.7,没有达到定量分析的要求;11种指标中儿茶素的相关系数最高为98.21,EGC的相关系数最低为56.7。研究结果为茶叶品质属性的快速准确甄别提供了一种新方法。更多还原

【Abstract】 Near infrared spectrum (NIRS) is a kind of electromagnetic wave between visible spectrum and middle infrared spectrum. Recently, NIRS has been developed as a kind of novel analysis and research method due to its characteristics of quick, convenient and relatively accurate.It is a major problem in current tea business to develop techniques for the fidelity identification of tea quality in a short time. Nowadays, fidelity identification of tea quality is generally performed by methods of sensory evaluation and fingerprint identification (including biological fingerprint and chemical fingerprint). However, both of these two methods have limitations. NIRS, as a newly developed method for analysis and research, has its advantages in application of fidelity identification of tea quality. Depend on the acquired complete spectrum information of tea samples, construct mathematical models by mathematical methods which can fully represent the characteristics of whole spectrum and exclusive attributes of tea samples. It would be able to find out the exclusive attributes of tea samples covered by lots of complex common factors. What is more, it would possibly be able to depict the resemblances among spectrums form different species of tea samples qualitatively or quantitatively. That is the reason why different tea samples could be categorized quickly and objectively.In this thesis, 66 tea samples were employed for the nondestructive identification category determination by means of NIRS analysis techniques combing with different chemical metrological methods (partial least-square method, standard method and factorization method). The results demonstrated that it was feasible to categorize Huangshan maofeng green tea samples accurately from other tea samples with different producing areas, species and processing techniques based on the NIRS information of the tea samples. As the combined chemical metrological testing methods, the standard method and factorization method was generally better than partial least-square method. Based on the chemical analysis data of 11 major components for the 66 tea samples by national approved standard methods, the quantitative analysis models were developed by the method of NIRS. During the process of model developing, 11 kinds of spectrum pretreatment methods are employed, including constant offset elimination,first derivative, second derivative, vector normalization, straight line subtraction, min-max normalization, multiplicative scattering correction (MSC), first derivative + straight line subtraction, first derivative + vector normalization, first derivative + multiplicative scattering correction (MSC) and no spectral data preprocessing. Seventy percent of samples were picked at random to construct models, then the left 30% of samples were used for models externally test. The results indicated R2 of moisture content, water extracts content, fiber content, caffeine, ECG, EGCG and catechin reached to 94.73, 93.52, 95.37, 97.39, 96.73, 95.24 and 98.21, respectively. And the R2 of C and EC were 84.89 and 86.34, which had met the requirements of quantitative analysis. While the R2 of ash content and EGC were only 71.61 and 56.7, which had not met the requirements of quantitative analysis. The highest R2 of the 11 major components is catechin, which was 98.21. And the lowest R2 was EGC, which was 56.7. The results provided a novel method for category and attribute determination of tea quality accurately and quickly.更多还原