论文部分内容阅读
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。
Near-infrared quantitative analysis of iodine value of palm oil was carried out by using visible-near-infrared transmission spectroscopy and CARS variable optimization method. By comparing the modeling results produced by different pretreatment methods, an ideal pretreatment method was found. Through the CARS variable selection method, a total of 60 effective wave points related to the iodine value of palm oil were optimized, and 60 valid wave points Establishment of palm oil iodine value optimization model. The RMSE (RM SEC = 0.0398) and the root mean square error of prediction (RM SEP = 0.0406) were obtained according to the set of model correlation coefficient (R_c = 0.9814) and the prediction set correlation coefficient (R_p = 0.9806) ) Is better than the coefficient error obtained by using the model established in the whole band. Using the visible near-infrared transmission spectroscopy coupled with CARS variable optimization method, the palm oil iodine value model is simplified and the accuracy of the iodine value prediction can be guaranteed.