若干中药的太赫兹时域光谱探测与分层聚类

Terathertz Spectroscopy of Chinese Traditional Medicines and Hierarchical Clustering Analysis

采用太赫兹时域光谱系统对3组相似中药进行了测量。经过傅里叶变换后,得到了6种中药的频域曲线,计算得到了它们的吸收系数作为分层聚类方法的特征数据,采用组间连接的聚类方法,以欧式平方距离作为度量标准,对中药进行了分层聚类分析与识别,聚类准确率达到了94.4%。太赫兹时域光谱技术为今后中药的鉴定与分类提供了一种有效的方法。

Several Chinese traditional medicines had been measured by terahertz time domain spectroscopy system. After Fourier transform, their frequency domain curve and absorption spectra were obtained. Clustering method of connection between the two groups and European square distance was applied to analyse the terahertz time domain spectroscopy of Chinese traditional medicine. Hierarchical clustering methods can be effective on classify the terahertz time domain spectroscopy of Chinese traditional medicine,as a clustering accuracy of 94.4 %. Terahertz time-domain spectroscopy provides an effective method for identification and classification of Chinese medicine in the future.