提高近红外吸收光谱法复杂溶液组分浓度分析精度的研究

发布时间：2018-02-25 21:27

本文关键词： 复杂溶液近红外吸收光谱检测精度光谱重叠系数等效信噪比光谱平方项多光程光谱　出处：《天津大学》2014年博士论文　论文类型：学位论文

【摘要】：为了对复杂溶液中被测组分的近红外吸收光谱法定量分析精度进行预估和提高，本研究系统地分析了近红外光谱法对复杂溶液组分进行检测的数学模型，分析了误差的来源，研究了复杂溶液样品本身特性和测量分析过程引入的各种因素对近红外吸收光谱法组分浓度定量分析精度的影响，提出多种提高近红外光谱定量分析精度的方法，并通过血清中总蛋白含量检测验证了优化方法的适用范围和有效性。主要研究工作如下： 1、从待测样品各组分的含量和光学特性的角度出发，对待测样品各组分浓度、消光系数曲线的形状和重叠程度，以及吸收系数和散射系数的大小对近红外光谱定量分析精度的影响进行分析。提出了一种与浓度分析精度有确定关系的重叠系数，对消光系数曲线的重叠程度进行度量。通过对乙醇水溶液浓度的光谱定量分析，验证了重叠系数与浓度分析精度的定量关系。 2、从光谱分析中测量条件的角度出发，研究光谱建模分析中光谱仪器噪声和参与建模的波长数两个因素与定量分析精度之间的关系，提出等效信噪比的概念。讨论了相同等效信噪比波段和不同信噪比波段中多波长参与建模对分析精度的影响。 3、研究了测量中被测组分和背景组分的浓度分布对被测组分光谱定量分析建模和预测精度的影响。分别分析了纯吸收被测组分、纯吸收背景组分、含散射被测组分和含散射背景组分在建模集和预测集中的不同浓度分布对光谱定量分析精度的影响。论文还讨论了参与建模的样品数量、样品池厚度等光谱分析测量条件对样品中被测组分浓度的定量分析精度的影响。 4、从获取溶液的散射信息并充分利用组分的散射特性的角度，提出使用利用平方项及高次项光谱、多光程光谱和多点分布式光谱提高光谱定量分析精度的检测方法。分析不同吸收和散射条件下各方法对被测组分浓度定量分析精度的影响。通过对血清中总蛋白浓度的定量分析，验证了建模中引入光谱平分项和高次项以及使用多光程光谱同时参与建模均可提高溶液组分定量分析精度。
[Abstract]:In order to estimate and improve the quantitative analysis accuracy of the components in complex solution by near - infrared absorption spectroscopy , this paper systematically analyzes the mathematical model for the detection of complex solution components by near - infrared spectroscopy , analyzes the source of error , studies the influence of various factors introduced by the characteristic of complex solution sample and the analysis process of measurement and analysis , and puts forward a variety of methods to improve the accuracy of quantitative analysis of near - infrared spectrum . The main research work is as follows : 1 . Based on the content and the optical characteristics of each component of the sample to be measured , the influence of the concentration of each component of the sample to be measured , the shape and the degree of overlap , and the absorption coefficient and the scattering coefficient on the accuracy of the quantitative analysis of the near infrared spectrum are analyzed . A quantitative relationship between the overlap coefficient and the concentration analysis accuracy is verified by the quantitative analysis of the concentration of the ethanol aqueous solution . 2 . Based on the measurement conditions in the spectral analysis , the relationship between the spectral instrument noise and the wavelength numbers involved in the modeling analysis and the quantitative analysis accuracy are studied . The concept of equivalent signal - to - noise ratio is proposed . The influence of the multi - wavelength participation modeling on the analytical precision in the same equivalent signal - to - noise ratio band and different signal - to - noise ratio bands is discussed . 3 . The influence of the concentration distribution of the tested components and background components on the quantitative analysis model and prediction accuracy of the measured components was studied . The effects of different concentration distributions on the spectral analysis accuracy of the components , pure absorption background components , the scattering - tested components and the scattering background components in the modeling set and the prediction set were analyzed respectively . The influence of the sample quantity , the thickness of the sample cell and the like on the quantitative analysis accuracy of the measured component concentration in the sample was also discussed . 4 . By analyzing the scattering information of the solution and making full use of the scattering characteristics of the components , this paper presents a method for detecting the accuracy of the quantitative analysis of the concentration of the components under different absorption and scattering conditions . The quantitative analysis of the concentration of the total protein in the serum is analyzed . By the quantitative analysis of the total protein concentration in the serum , it is verified that the introduction of the spectral bisection item and the high order term in the modeling and the simultaneous participation in the modeling using the multi - optical path spectrum can improve the quantitative analysis accuracy of the solution component .

【学位授予单位】：天津大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：R917;O657.33

【相似文献】