[关键词]
[摘要]
近年来我国中药产业发展迅速,但中药质量控制的方法仍不全面,是阻碍其发展的主要因素。模式识别法于20世纪80年代引入化学研究领域,同时也被应用于中药研究领域,目前,许多学者以模式识别理论为基础,建立了多种中药的有效且科学的质量评价方法。通过对模式识别法的基本原理和方法及其几种分析技术,如主成分分析、聚类分析、判别分析、灰色关联分析、偏最小二乘法、直观推导式演进特征投影法和人工神经网络技术等在中药质量控制及评价方面的应用进行综述,以期为模式识别法在中药质量评价中的进一步应用提供参考。
[Key word]
[Abstract]
In recent years,with the development of Chinese materia medica (CMM) industry,the problem of quality control method is not comprehensive.It becomes the most important factor to block the development of CMM.Many scholars look for some new methods,by which the quality of CMM could be assessed objectively and accurately.Since 1980s,the pattern recognition was introduced to the chemical field and was applied to CMM at the same time.And there were some CMM quality assessment methods based on the pattern recognition established so far.In this article,the latest research progress in the pattern recognition,basic principle,and technology,was introduced and the applications in principal component analysis,cluster analysis,discriminant analysis,grey correlation analysis,partial least squares,heuristic evolving latent projections and artificial neural networks of CMM quality assessment have been reviewed,so as to provide the reference for further studies and application in CMM.
[中图分类号]
[基金项目]
吉林省科技计划发展项目(20140204063YY)