摘要
已知组成范围的“灰色”分析体系的校正,其首要步骤是鉴定出体系中实际存在的物种.本文对由所有可能存在组分的标准光谱与样品光谱构成的矩阵进行了奇异值分解,根据对应奇异向量的元素判断样品的组成,继而做定量分析.计算机模拟与实际分析表明方法令人满意.
The first step of the calibration oi 'grey' analytical systems with known scope of pos-sibiy existing components is So identify the truely existing species in a specific sample.In this pa-per,a specially designed matrix which consists of the spectra of all possibly existing species and the analytical sample was constructed,and a method based on the singular value decomposition (SVD) of this matrix was developed to pick out the really components existing in the mixture according to the elements of the relevant singular vector The qualitative detection was followed by quantitative determination by using ordinary regression procedure The method was testified by simulated spectra and experimentally measured data with satisfactory results.
出处
《高等学校化学学报》
SCIE
EI
CAS
CSCD
北大核心
1992年第8期1033-1038,共6页
Chemical Journal of Chinese Universities
基金
国家自然科学基金
关键词
奇异值分解
多元校正
灰色体系
Singular value decomposition(SVD) ,Muitivate calibration,'Grey' analytical system,Chemornetrics