龙泉窑古陶瓷年代断定的中子活化分析和Bayes判别  

Dating Ancient Chinese Celadon Porcelain by Neutron Activation Analysis and Bayesian Classification

作　　者：谢国喜[1,2] 冯松林[1] 冯向前[1] 朱继浩[1] 闫灵通[1] 李丽[1] 

机构地区：[1]中国科学院高能物理研究所,北京100049 [2]中国科学院研究生院,北京100049

出　　处：《原子能科学技术》2009年第6期561-565,共5页Atomic Energy Science and Technology

基　　金：中国科学院知识创新工程资助项目(KJCX3.SYW.N12);国家自然科学基金资助项目(50432010,10675143);北京市自然科学基金资助项目(1082009);中国科学院高能物理研究所青年创新基金资助项目(542007IHEPZZBS546100)

摘　　要：为实现龙泉窑古陶瓷年代的准确判定,利用中子活化分析(NAA)方法准确测定了南宋-元(AD 1127—1368)和明代(AD 1368—1644)龙泉窑古瓷胎中La、Sm、U、Ce等18种元素的含量。为提高实验数据的可靠性,利用马氏距离准则排除实验数据的异常值,并对实验数据进行了正态分布检验。在一定的假设前提下,应用Bayes判别法用预处理的实验数据对样品进行年代判别。结果表明:明代69个古陶瓷样品全部正确归类,南宋-元时期96个古陶瓷样品仅有3个未能正确归类,全部样品归类的正确率为98.2%,说明NAA技术和Bayes判别法可望在古陶瓷的年代断定研究中发挥重要作用。Dating ancient Chinese porcelain is one of the most important and difficult problems in porcelain archaeological field. Eighteen elements in bodies of ancient celadon porcelains fired in Southern Song to Yuan period （AD 1127-1368） and Ming dynasty （AD 1368-1644）, including La, Sm, U, Ce, etc. , were determined by neutron activation analysis （NAA）. After the outliers of experimental data were excluded and multivariate normal distribution was tested, and Bayesian classification was used for dating of 165 ancient celadon porcelain samples. The results show that 98.2% of total ancient celadon porcelain samples are classified correctly. It means that NAA and Bayesian classification are very useful for dating ancient porcelain.

关 键 词：年代断定 古陶瓷 中子活化分析 Bayes判别法 

分 类 号：O434.19[机械工程—光学工程] O657.34[理学—光学]