基于近红外三波长检测葡萄糖水溶液浓度的两种预测模型比较

The Comparison of Two Prediction Models Based on Near-Infrared Three Wavelengths of Glucose Concentrations Measurement in Aqueous Glucose Solutions

讨论了近红外三波长测定水溶液中葡萄糖浓度的建模方法,采用偏最小二乘回归建立预测模型,对预测集样品中的葡萄糖浓度进行预测,结果所得预测均方根误差(RMSEP)为17.08mg/dL,浓度预测值与参考值的相关系数为0.997,而采用吸光度差值比率法所得预测模型的RMSEP达到54.94mg/dL,相应的相关系数仅为0.949。结果表明偏最小二乘回归所建模型用于近红外三波长葡萄糖浓度的预测,所得预测误差较小,初步验证了系统的可行性。

The modeling methods was discussed to determine glucose concentrations based on near-infared three wavelengths in aqueous glucose solutions in the text. The model was established based on partial least-square algorithm. When it was used to predict the glucose contents of validation set, the root mean square error of prediction （RMSEP） equaled 17.08mg/dL and the correlation coefficient between prediction values and reference values attained 0.997 ; while using logarithmic ratio of absorbency difference method to set up prediction model, the RMSEP reached 54.94mg/dL and the correlation coefficient equaled 0.949 merely. The results suggest that the model based on PLS algorithm has lesser RMSEP, it also validates the system＇s feasibility primarily.