在XRF多变元体系因素分析中应用交互有效法选择体系的主组份  

THE USE OF CROSS-VALLDATION TO ESTIMATE THE NUMBER OF PRINCIPAL COMPONENT IN XRF MULTIVARIATION CALIBRATION SYSTEM

作　　者：罗立强[1] 梁国立[1] 马光祖 

机构地区：[1]中国地质科学院岩矿测试技术研究所,北京100037

出　　处：《光谱学与光谱分析》1993年第5期113-118,26,共7页Spectroscopy and Spectral Analysis

摘　　要：本文报道了在X-射线荧光分析多变元体系因素分析中,应用交互有效法选择体系主组份的研究。本方法在基本参数法的基础上,运用交互有效法,通过最小预报残差平方和(PRESS)的计算,对体系进行了主组份分析,确定了该体系中的最佳维数和主要影响因素。选择体系中的主组份进行回归分析,剔去了噪声,模型稳定性增强,XRFA的预报能力和分析结果准确度有了显著的提高。文中给出了分析地质样品中十三个主、次量元素的实验结果和数据对照。Cross-validation method is used to estimate the number of principal component in X-ray fluorescence analysis. Based on fundamental parameters method, the number of factors in XRF multi-variation calibration system has been estimated through the calculation of PRESS (predictive residual error sum of squares) and the analysis of principal component. By using cross-validation with fundamental parameters method, a noise filtering effect is obtained, the stability of XRF calibration models is enhanced, and the prediction ability and analytical accuracy improved obviously. The analytical results of 13 elements in the geoIogi cal samples are reported here. The program is written in Fortran 77 runs and in UNIX, MSDOS and XENIX operation systems.

关 键 词：多变元体系 因素分析 荧光分析 

分 类 号：O657.34[理学—分析化学]