近红外漫反射光谱法快速检测胶原肽粉中低聚肽含量

Rapid determination of oligopeptides content in collagen peptide powder by near infrared reflectance spectroscopy

目的建立近红外漫反射光谱法(near infrared reflectance spectroscopy,NIRS)快速检测胶原肽粉核心功效成分低聚肽含量的分析方法。方法基于径向基神经网络(radial basis function neural network,RBFNN)的方法,分析比较了多元散射校正(multiple scatter calibration,MSC)、标准正态变换(standard normal variation,SNV)的光谱预处理方法,分别结合了主成分分析(principal componentanalysis,PCA)、独立成分分析(independent component analysis,ICA)的方法提取特征信息,优化输入变量、扩展系数等参数,建立了低聚肽含量的检测模型,同时为了比较方法优劣,也建立了相应的偏最小二乘方法(partial least squares,PLS)模型。结果非线性建模方法RBFNN比线性PLS方法模型效果更好,与常用的PCA-RBFNN模型相比,本研究采用的ICA-RBFNN模型结果更准确,模型独立验证集的相关系数R是0.87,预测标准偏差(root mean square error of predicition,RMSEP)是1.71%。结论所建立的模型准确度高,适用于胶原肽粉低聚肽含量的快速高效分析。

Objective To establish a method for the rapid determination of the core functional components oligopeptides content of collagen peptide powder by near infrared reflectance spectroscopy(NIRS).Methods Radial basis function neural network(RBFNN)was adopted to establish the NIRS detection models,the spectral preprocessing methods were compared between multiple scatter calibration(MSC)and standard normal variate(SNV).Principal component analysis(PCA)and independent component analysis(ICA)were combined to extract feature information,optimize input variables,expansion coefficient and other parameters,and establish the detection model of oligopeptide content.At the same time,in order to compare the advantages and disadvantages of the method,the corresponding partial least squares(PLS)model was also established.Results The nonlinear RBFNN models were better than the essentially linear PLS model and the proposed ICA-RBFNN was more efficient than the conventional PCA-RBFNN.the correlation coefficient(R)and the root mean square error of predicition(RMSEP)of the model was 0.87 and 1.71%,respectively.Conclusion The established model is accurate and the NIRS method is promising for determination the oligopeptides of collagen peptide powder quickly and efficiently.