| 引用本文: | 宋江婷,金福江,周丽春.透射光谱线性空间核学习建模求解多组分浓度[J].控制理论与应用,2024,41(3):468~473.[点击复制] |
| SONG Jiang-ting,JIN Fu-jiang,ZHOU Li-chun.Kernel learning modeling for solving multi-component concentrations with linear space of transmission spectra[J].Control Theory and Technology,2024,41(3):468~473.[点击复制] |
|
| 透射光谱线性空间核学习建模求解多组分浓度 |
| Kernel learning modeling for solving multi-component concentrations with linear space of transmission spectra |
| 摘要点击 2879 全文点击 248 投稿时间：2022-09-30 修订日期：2023-05-15 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/CAT.2023.20863 |
| 2024,41(3):468-473 |
| 中文关键词 多组分 浓度 量子遂穿 线性系统 |
| 英文关键词 multi-component system concentration quantum Tunneling linear systems |
| 基金项目 福建省科技计划项目(2021H6028), 流程工业综合自动化国家重点实验室联合开放基金项目(2022–KF–21–04)资助. |
|
| 中文摘要 |
| 本文对求解多组分体系浓度的分光光度计同时测定法提供了一种建模方法. 通过对多组分体系的透射光
谱建立光谱线性空间, 证明了多组分体系的分子势函数矩阵VM可以转化为对角矩阵, 对角矩阵的对角元是单组分
势函数的块矩阵Ji. 多组分体系透射光谱空间是单组分分子透射光谱函数为基函数、分子数量占比为坐标的线性
组合. 利用多核学习法确定各单组分体系透射波的占比权重系数, 提出了用单组分浓度光量子隧穿软测量模型测定
多组分浓度的测量方法. 实例验证表明此方法稳定可靠, 且能降低计算的复杂性. |
| 英文摘要 |
| A modeling method is proposed in this paper for simultaneous spectrophotometric determination of multicomponent
system concentrations. A linear space of transmission spectra of the multi-component system was established,
which proved that the potential function matrix VM of the multi-component system can be transformed into a diagonal
matrix, in which the diagonal elements are blocked matrix Ji of the single-component potential functions. The multicomponent
system transmission spectral space is a linear combination of the transmission spectral functions of each singlecomponent
molecule as the basis function and the proportion of molecular numbers as the coordinate. The weight coefficients
of the transmission waves of each single-component system were determined using the multi-kernel learning method,
and a measurement method for measuring multi-component concentration using a single-component concentration quantum
tunneling soft measurement model was proposed. Example validation shows that this method is stable and reliable and
can reduce computational complexity. |