首页|基于PCA-SVR算法和共聚焦拉曼光谱的牡丹籽油掺伪定量分析研究

基于PCA-SVR算法和共聚焦拉曼光谱的牡丹籽油掺伪定量分析研究

扫码查看
牡丹籽油掺伪现象严重,需要加强监管和打击力度,如何鉴别牡丹籽油的真伪成为了一个亟待解决的问题.该研究旨在探究基于PCA-SVR算法和共聚焦拉曼光谱的牡丹籽油掺伪定量分析方法.首先,采集牡丹籽油样品,并进行共聚焦拉曼光谱分析以确定其化学成分和结构,分别测试不同掺伪浓度的牡丹籽油的混合油.然后,利用拉曼光谱数据建立PCA-SVR定量分析模型,以区分牡丹籽油和其他植物油的掺伪情况.最后,通过实验验证该方法的准确性和可行性,结果表明该方法可以有效地鉴别牡丹籽油和其他植物油的掺伪情况,模型测试集量化效果的决定系数R2优于0.98,均方根误差RMSE小于0.04.该方案检测牡丹籽油掺伪浓度的理论极限为0.011 52,即对于预测混合油品中牡丹籽油所含体积比大于0.011 52的样品是可信的.基于PCA-SVR算法和共聚焦拉曼光谱相结合的检测方案具有较高的准确性和可靠性,该研究对于提高牡丹籽油的质量控制和保障消费者的健康具有重要意义.
Quantitative Analysis of Adulteration of Peony Seed Oil based on PCA-SVR and Confocal Raman Spectroscopy
It is imperative to implement more stringent punitive measures to combat the adulteration of peony seed oil,and establish quantitative methods to determine the authenticity of peony seed oil.A quantitative analysis model,based on principal component analysis and support vector regression(PCA-SVR)and Raman spectroscopy was developed,to determine the chemical composition and structure of samples of peony seed oil mixed with different proportions of soybean oil.The method effectively identified adulteration of peony seed oil.The coefficient of determination(R2)of the quantitative effect of the model test set exceeded 0.98 and the root-mean-square error(RMSE)was less than 0.04.The theoretical limit for detecting the adulteration in peony seed oil in this model was 0.011 52,indicating that the method was reliable for measuring mixed oil samples with a volume ratio of peony seed oil that was greater than 0.011 52.The detection model combining PCA-SVR and confocal Raman spectroscopy demonstrated high accuracy and reliability.The findings of this study have important implications for improving the quality of peony seed oil to ensure consumer health.

principal component analysis and support vector regressionconfocal Raman spectroscopypeony seed oilquantitative analysis

周会国、赵辉

展开 >

广西职业技术学院大数据学院,广西南宁 530226

河北工程大学信息与电气工程学院,河北邯郸 056038

PCA-SVR算法 共聚焦拉曼光谱 牡丹籽油 定量分析

2024

现代食品科技
华南理工大学

现代食品科技

CSTPCD北大核心
影响因子:1.07
ISSN:1673-9078
年,卷(期):2024.40(11)