可见-近红外光谱结合CARS优化模型预测棕榈油碘值

Visible near infrared spectroscopy combined with CARS iodine value optimization model for prediction of palm oil

采用可见-近红外透射光谱结合CARS变量优选方法优化模型,对棕榈油碘值进行近红外定量分析。通过将使用不同预处理方法产生的建模效果进行比较,找到了理想的预处理方法,通过CARS变量选择方法优选出与棕榈油碘值相关的有效波点共60个,利用60个有效波点建立棕榈油碘值优化模型。根据优化模型的建模集相关系数(R_c=0.9814)和预测集相关系数(R_p=0.9806),得到的建模均方根误差(RM SEC=0.0398)和预测均方根误差(RM SEP=0.0406)优于采用全波段建立的模型得到的系数误差。利用可见近红外透射光谱结合CARS变量优选方法,简化了棕榈油碘值模型,并能够保证碘值预测的准确度。

In this paper, optimized by near-infrared transmission spectroscopy combined with CARS optimization method, the value of palm oil for iodine （I）was quantitatively analyzed by near infrared method. Comparing the fitting results produced by different pretreatment methods, the ideal pretreatment method was found out. By CARS variable selection method, 60 effective points related to iodine value of palm oil were selected, and with these points, the optimal model of the iodine value of palm oilwas established. The root mean square error of cross validation（ RMSEC = 0. 0398） and the root-mean-square error of prediction（ RMSEP = 0. 0406） are better than those obtained from the whole model, Which is set.based on the correlation coefficient （Re = 0. 9814） and the prediction set correlation coefficient （ Rp = 0. 9806） of the optimization model. By using the visible near-infrared transmittance spectroscopy combined with the CARS variable optimization method, the iodine value model of palm oil has been successfully and effectively simplified, and the accuracy of iodine value prediction can be ensured.