聚类分析法在快速鉴别中药牵正散中的应用

Rapid and undamaged determination of qianzheng power by fourier-transform infrared spectroscopy and clustering analysis

以牵正散及其不同配伍组分的近红外图谱作为聚类分析的对象,在建立混合成分模型的基础上,采用SIMCA聚类分析法快速分类研究牵正散及其不同配伍组分近红外光谱特征。实验中,分别设牵正散全方、缺白附子、缺全蝎和缺僵蚕的4个样品。结果表明:由于牵正散及其不同配伍组分的近红外图谱变异度较大,其光谱聚类的结果较理想,盲样检测的正确率可达90%。因此,如果足够的样本,增加训练集样本数和采样的代表性,加强操作的标准,其准确率将大大提高。所以,近红外光谱与聚类分析相结合的方法可以快速、无损识别复方中药。

On the basis of building a model of main components and with the fingerprint infrared spectrum as the object, Qianzheng Powders with different alignments respectively were rapidly classified and closely studied by using the method of clustering analysis. Because of the small difference between the spectra of different compound recipes, including the compound recipe, the compound recipe without Rhizoma Typhonii, the compound recipe without Scorpio and the compound recipe without Bombyx Batryticatus, the outcome of the clustering is ideal. When testing with the blind samples picked out from the chosen samples, the accuracy of clustering reaches up to 90%. In order to increase the clustering accuracy, the number of the samples should be increased; additionally these samples should be more characteristic. On the whole, with the combination of the fingerprint spectrum and the clustering analysis, the compound recipe can be distinguished rapidly.