人参作为东北地区重要的经济作物, 是著名的“东北三宝”之一。 随着社会经济的快速发展, 人们对于养生保健产品的重视度和需求量也随之提高。 目前, 市场所售人参存在农残超标、 品质较差和种类混乱的现象, 而且含有重金属元素的人参对于食用者会造成极大的危害。 传统人参识别方法主要是根据人参基源、 性状、 显微和理化特性对人参进行生药研究, 但是这些方法存在人为因素影响、 繁琐复杂的样品预处理和二次污染的问题, 不能实现可靠快速检测。 基于激光诱导击穿光谱技术（LIBS）结合主成分分析法（PCA）对人参产地和人参部位进行聚类分析, 以及对人参重金属元素进行定量分析。 采集吉林省5个产地6种人参的LIBS光谱数据（200~975 nm）, 通过对光谱进行平均值预处理, 利用PCA算法对光谱数据进行降维、 聚类分析, 实现人参产地快速聚类识别、 同一产地的园参与林下参的聚类识别以及同一株人参的不同部位聚类识别。 实验结果发现, 由于同株人参不同部位的元素含量区别不大, 造成人参主干和参须部位聚类效果不佳, 对于6种人参可实现较好的分类效果。 最后对人参样品中重金属元素Pb和Cr进行了定量分析, 计算得到人参中Pb和Cr元素的检测限（LOD）分别为9.55和10.86 mg·kg-1, 去一交互均方根误差（RMSECV）分别为0.011和0.023 Wt.%。 结果表明LIBS结合PCA算法对于人参分类和重金属检测具有较好的应用前景。

The Panax ginseng is one of the most important commercial crop and precious medicine in northeastern China. With the rapid development of the economy, people’s living standard continuously improved, so the demand for health products also increased levels. At present, Due to the market mechanism has less of management and supervision measures, some problems of excessive pesticide residues, poor quality, confusion of quality and variety need to be solved. The heavy metal in Panax ginseng is extremely harmful to human health. The traditional analysis method of Panax ginseng classification is mainly based on the ginseng origin, shape, microscope and physicochemical properties. However, these methods have some problems, such as human factors, complicated sample pretreatment and secondary pollution, and reliable and rapid detection is not possible. In this study, laser-induced breakdown spectroscopy (LIBS) combined with the principal component analysis (PCA) algorithm model is established. A set of six habitats from five different locations, and three different part of Panax ginseng samples were used for LIBS experiment, the mean values of the LIBS spectrum (200~975 nm) were pretreated. The experiment results found that by dimensionality reduction and cluster analysis of spectral data, the PCA models have a good ability of classification for different habitats Panax ginseng the six habitates ginseng has a better classification. Finally, the quantitatively analyzed method was proposed, the limits of detection (LODs) of Pb and Cr is 9.55 and 10.86 mg·kg-1, the RMSECV is 0.011 Wt.% and 0.023 Wt.%, respectively. It is shown that LIBS combined with PCA algorithm to the ginseng classification and heavy metal detection has good effect and foreground.