晶体的X射线衍射物相分析方法研究  

Study on phase analysis method of X-ray diffraction of crystals

作　　者：王新[1] 徐捷[1] 穆宝忠[1] WANG Xin;XU Jie;MU Baozhong(School of Physics Science and Engineering,Tongji University,Shanghai 200092,China)

机构地区：[1]同济大学物理科学与工程学院,上海200092

出　　处：《实验技术与管理》2021年第3期29-33,共5页Experimental Technology and Management

基　　金：国家重点研发计划项目(2017YFA0403304)。

摘　　要：该文提出了一种基于多元判断分析算法的晶体物相分析模型,开展了晶体的X射线衍射物相分析研究。模型利用主成分分析算法提取晶体衍射谱的特征信息,结合光谱角匹配和相关系数算法对衍射谱的轮廓和相关度进行分析。实验测量了单晶和多晶物质的X射线衍射谱,构建了样本数据库,利用模型对实测衍射谱进行了物相分析。样本的训练结果表明,与单一的物相分析算法相比,多元判断分析算法对实测的衍射谱具有更高的识别精度,相似度最接近的两种晶体的SD值可达0.8,实现了对晶体物相类别的精确识别。A model of phase analysis of crystals based on multivariate discriminant analysis algorithm is proposed, and the X-ray diffraction phase analysis of crystals is carried out. The model utilizes the principal component analysis algorithm to extract the characteristic information of the diffracted spectrum of crystals and combines the spectral angle matching and correlation coefficient algorithm to analyze the contour and correlation degree of the diffracted spectrum. X-ray diffracted spectra of monocrystal and polycrystalline materials are measured experimentally, and the sample database is constructed. The measured diffraction spectra are analyzed by the model. The results of training show that compared with the single-phase analysis algorithm, the multivariate discriminant analysis algorithm has higher identification accuracy for the measured diffraction spectra, and the SD value of the two types crystals with the closest similarity can achieve 0.8, which realizes the accurate identification for the phase categories of crystals.

关 键 词：X射线衍射 物相分析 多元判断分析 晶体 

分 类 号：O434.1[机械工程—光学工程]