摘要
将流动注射在线预富集技术引入光度分析,较好地解决了一般树脂相光度法重现性不好的问题。提出了应用淋洗曲线作为多组份同时测定的定量依据;采用人工神经网络法对体系的非线性数据进行处理,较好地解决了复杂的地质样品中痕量贵金属元素Au、Pd、Ag的同时测定。并与PLS、NPLS多元校正方法进行了对比,显示出人工神经网络法具有高度非线性表达能力。本文还对B-P算法的应用及其影响因素作了较深入的探讨。
A new system of flow injection on-line preconcentration with VS- I anion exchange fiber was established for simultaneous spectrophotometric determination of trace Au, Pd and Ag in geological samples. A FIA simultaneous determinaion method based on elution curves was brought up. The ANN calculation was used to deal with the nonlinear data, and the application and influence factors of B-P algorithm were investigated. The study on three kinds of multivariate calibration methods showed that the B-P algorithm has a strong correction capability for nonlinear data.
出处
《分析化学》
SCIE
EI
CAS
CSCD
北大核心
1998年第1期7-11,共5页
Chinese Journal of Analytical Chemistry
基金
国家自然科学基金资助课题。
关键词
光度分析
流动注射
ANN
金
钯
银
地质样品
Artificial neural network, multicomponent, spectrophotometry, flow injection, noble metal