采用激光诱导击穿光谱（LIBS）技术结合主成分分析（PCA）, 对凌家滩遗址出土陶器进行快速检测和统计分类的研究。 凌家滩遗址位于安徽马鞍山市含山县, 是我国南方地区一处大型新石器时代晚期聚落, 遗址出土大量精美玉器、 石器和陶器等, 年代处于中华文明起源关键节点上, 是研究我国文明起源和发展问题的重要遗址, 因此对其出土陶器的研究具有十分重要的文化历史意义。 对凌家滩遗址出土具有不同类型羼和料的陶器样本, 进行LIBS快速分析后, 利用PCA对结果进行统计辨析。 研究表明, 不同羼和料的光谱敏感度有所不同, 不同掺杂物质会产生光谱特征差异。 另一方面, 出于统计分析的考虑, 有目的缩小了背景噪声等异常数据干扰, 并进行了分类辨析, 在元素谱线归属的基础上进行了特征谱线提取, 实现了利用多元统计分析进行快速分类的目的。 分析结果表明, 与泥质陶相比, 羼植物陶器和羼细砂类样品, LIBS光谱特征具有良好的辨识度, 可以进行有效区分。 其他类型羼和料根据实际情况不同, 则需要配合其他手段进行综合判断。 相关研究结果可对南方地区新石器时期陶器的快速鉴定和类型归属等工作, 提供了科学支持和有益借鉴。

For applying the Laser Induced Breakdown Spectroscopy technology to the ancient pottery(Lingjiatansite, 3000BC, Anhui, China) research, the goal and aim is for quick identificationand classification of the different types of ancient ceramic wares. Lingjiatan Site, located in Hanshan County, Maanshan City, Anhui Province, China, is a large late Neolithic settlement in southern China. A large number of fine jade articles, stone ware and pottery have been unearthed from the site. It is an important site!for studying the origin of earlyChinese civilization. Therefore, the study of its pottery is of great cultural and historical significance. After LIBS analysis, using the principal component analysis to process the data and give the reference to the classification workof pottery. The results show that different temper in body of pottery will affect the characters of spectrum and the PCA could give the classification group based on those spectra discrepancies. In the other hand, due to the consideration of statistical analysis, the abnormal data interference such as background noise is purposefully reduced and classified, and the characteristic spectral lines are extracted based on the attribution of the element spectral lines, thus realizing the purpose of rapid classification by multivariate statistical analysis. The results show that compared with pure argillaceous pottery, the samples of temper of plant pottery and some fine sand temper have well discrimination in LIBS spectral characteristics and can be effectively distinguished. According to the actual situation, other types of processing and materials need to be judged comprehensively with other means. This work indicates that the LIBS and PCA will be suitable and useful tools for ancient ceramics research.