摘要
由于光谱具备信噪比低、吸收强度较差等特性,导致不同温度下检测结果有较大差异。以葡萄糖溶液为例,探究温度对近红外光谱检测结果影响。研究近红外光谱检测原理,采用化学计量学法建立光谱信息与葡萄糖溶液浓度相关数学模型,对光谱、浓度矩阵进行线性回归分析,得到校准集模型回归系数矩阵;融合BP神经网络与主成分分析法对数据预处理,解决近红外光谱检测的非线性问题,剖析浑浊介质内葡萄糖近红外光谱,明确温度引起的光谱变异机制;设计纯水与葡萄糖水溶液光谱温度特性对比实验,明确不同温度对光谱的影响程度。由实验结果可知:在进行光谱检测时,应利用一定手段排除温度干扰,以此得到准确的检测结果。
Because the spectrum has the characteristics of low signal-to-noise ratio and poor absorption intensity,the detection results at different temperatures are quite different.This paper takes glucose solution as an example to explore the influence of temperature on the results of NIR.This paper studies the principle of near-infrared spectrum detection,establishes the mathematical model of correlation between spectrum information and concentration of glucose solution by chemometrics,carries out linear regression analysis of spectrum and concentration matrix,and obtains the regression coefficient matrix of calibration set model,integrates BP neural network and principal component analysis to preprocess the data,solves the nonlinear problem of near-infrared spectrum detection,and analyzes the near-infrared spectrum of glucose in turbid medium.The mechanism of spectral variation caused by temperature is determined.The comparative experiment of spectral temperature characteristics of pure water and glucose aqueous solution is designed to clarify the influence degree of different temperatures on the spectrum.From the experimental results,we can know that,in the spectral detection,we should use certain means to eliminate the temperature interference,so as to get accurate detection results.
作者
白心爱
刘艳娥
薛丽丽
BAI Xin’ai;LIU Yan’e;XUE Lili(Department of Physics Lvliang Collage,Lvliang Shanxi 033000,China)
出处
《激光杂志》
北大核心
2020年第10期97-101,共5页
Laser Journal
基金
山西省自然科学基金项目(No.201801D121125)
教育部高等教育司产学合作协同育人项目革(No.201802022055)
吕梁市重点研发计划(高新领域)。
关键词
近红外光谱
温度影响
光谱检测
葡萄糖浓度
near-infrared spectrum
temperature effect
spectral detection
glucose concentration
