近红外透射光谱分析油茶籽油掺入豆油的研究

采用傅里叶近红外透射光谱技术,应用于不同比例(0%～50%)的豆油和油茶籽油的二元体系样品,通过各种光谱预处理方法和回归方法的优化,建立了检测油茶籽油中掺杂豆油含量的近红外光谱的定量模型。以"二阶导数+Norris derivative filter"的最佳预处理方法和PLS的回归方法所建模型最佳,其校正相关系数(Rc)和校正标准误差(RMSEC)分别为0.999 99、0.057 70;交叉检验相关系数(Rcv)和交叉检验校准误差(RMSECV)分别为0.999 99、0.071 9;最优波段为5 037.16～4 728.60 cm-1、7 852.72～7 089.04 cm-1、8 577.82～8 323.26 cm-1;最佳主因子数为6。另外,经外部验证15份随机样品的化学值和NITS预测值的相关系数(R2)为0.998。表明近红外透射光谱法能够快速、准确地定量分析掺假油茶籽油中豆油的含量(范围为0%～50%)。

Study on the Quantitative Analysis of Camellia Oil Adulterated with Soybean Oil by Near Infrared Transmittance Spectroscopy(NITS)

The quantitative model of camellia oil adulterated with mixed different proportions(0%～50%)soybean oil was well developed by the near infrared transmittance spectroscopy with different spectral pretreatment methods and regression analysis.The results demonstrated that,the best model was established with second derivative and Norris derivative filter pretreatment methods and PLS regression analysis.Meanwhile,the correlation coefficients of calibration(Rc) and the root mean square errors of calibration(RMSEC) were 0.999 99 and 0.057 7;the correlation coefficients of cross-validation(Rcv) and the root mean square errors of cross-validation(RMSECV) were 0.999 99 and 0.071 9 respectively;the best spectral ranges were 5 037.16～4 728.60 cm-1,7 852.72～7 089.04 cm-1,8 577.82～8 323.26 cm-1;the best factor was 6.Furthermore,the correlation coefficients(R2) of chemical values and NITS values in external validation was 0.999 8.Therefore,the near infrared spectroscopy methods will broad application in detection of camellia oil adulterated with soybean oil(range from 0%～50%).