摘要
分析了溶质成分的浓度的改变通过吸收和散射效应对光谱的影响,利用通过浓度改变引起的吸收和散射效应相互抵消的波长作为基准波长,并使用基准波长法以有效的消除仪器漂移和温度变化等外界因素的干扰,从而提高模型的预测能力。通过处理脂肪乳溶液和血浆溶液中含不同葡萄糖的光谱数据,证明了基准波长法能提高模型的相关系数和降低模型预测均方差。
In the present paper, the authors analyzed the absorption and scattering effects of changing the solution concentration on the spectra, and dealed with the wavelength, where the absorption and scattering effects counteract completely, the fiducial wavelength method can eliminate the perturbation from shift of instrument and temperature, and also can improve predicting ability of the model. The result of utilizing this method to process the spectral data of two sample sets, i.e. intralipid solution and plasma solution with different level glucose, has verified that fiducial wavelength method can improve correlation coefficient and decrease RMSEP of predicting model.
出处
《光谱学与光谱分析》
SCIE
EI
CAS
CSCD
北大核心
2006年第8期1416-1418,共3页
Spectroscopy and Spectral Analysis
基金
国家十五攻关项目(2004BA706B12)
天津市自然科学重点基金(023800411)资助
关键词
混浊介质
基准波长法
相关系数
预测均方差
Turbid media
Fiducial wavelength method
Correlation coefficient
RMSEP