论文部分内容阅读
为了快速准确的鉴别续断发汗与否,以续断发汗和未发汗样品为实验材料,采用近红外光谱法结合主成分分析-马氏距离判别分析方法建立了定性鉴别模型。选取了129个未发汗样品和86个发汗样品的近红外光谱图,应用主成分分析-马氏距离法进行判别分析,选择谱段为9 881.46~4 119.20 cm-1,采用“标准正则变换+原始光谱+二阶求导”组合对原始光谱进行预处理,主成分数为14,建立定性鉴别模型;并经预测集验证,鉴别准确率达到100%。说明近红外光谱结合模式识别方法进行续断“发汗”与否定性鉴别在技术上是可行的,可以作为续断产地加工“发汗”定性鉴别的一种辅助手段。
In order to quickly and accurately identify the perspiration or not to perspire perspiration and non-sweating samples as experimental materials, the use of near infrared spectroscopy combined with principal components analysis - Mahalanobis discriminant analysis to establish a qualitative identification model. NIR spectra of 129 non-sweat samples and 86 sweat samples were selected and discriminant analysis was conducted by using principal component analysis-Mahalanobis distance method. The selected spectral bands were 9 881.46 ~ 4 119.20 cm-1, and the standard regular transformation + Original spectrum + second order derivative “, the original spectrum was preprocessed. The main component was 14, and a qualitative identification model was established. After the prediction set was verified, the accuracy of identification was 100%. It is technically feasible to use the method of near infrared spectroscopy in combination with pattern recognition to detect the ”sweating“ and negative identification, and it can be used as an aid to the qualitative identification of ”sweating" in the place of origin.