论文部分内容阅读
婴幼儿乳粉的质量和安全日益受到人们的关注,研究了利用近红外漫反射光谱进行乳粉中各类营养物质含量的快速无损检测的适应性。以156个乳粉样本作为样本集,分别采用偏最小二乘回归(PLSR)和KNN保形映射(KNN-KSR)算法建立含量范围分别为14.5%~23.1%(蛋白质、脂肪)、(5.6~9.2)mg/g(钙)、(1.29~10.2)mg/100 g(锌、V_(B2))、(0.34~1.47)mg/g(Vc)的营养物质近红外定量分析模型。PLSR与KNN-KSR预测各类营养物质含量的平均相对误差分别小于5%与3%,2种方法对Vc的预测误差最大。不同光谱预处理方法所得结果显示一阶导预处理结果更为理想,可使每种营养物质的预测平均相对误差降低0.1%以上。将样品按照Vc浓度的量级划分为2个区间分别建立模型,平均相对误差较整个浓度区间建模结果减小1%以上。利用PLSR与KNN-KSR方法基于近红外光谱信息可快速预测乳粉中不同浓度量级的营养成分,KNN-KSR的预测效果更佳。为获得准确的预测结果,建议采用同量级浓度样品建立乳粉营养成分的近红外定量模型并分浓度区间进行应用。
The quality and safety of infant milk powder are getting more and more attention. The adaptability of rapid non-destructive testing of various nutrients in milk powder by near-infrared diffuse reflectance spectroscopy was studied. A total of 156 samples of milk powder were used as sample sets. The contents of proteins were determined by partial least squares regression (PLSR) and KNN conformal mapping (KNN-KSR) respectively, and their contents ranged from 14.5% to 23.1% (protein, 9.2) mg / g (Ca), (1.29-10.2) mg / 100 g (Zn, V B2) and (0.34-1.47) mg / g (Vc) The average relative errors of PLSR and KNN-KSR in predicting the content of various nutrients were less than 5% and 3%, respectively. The prediction error of Vc was the highest among the two methods. The results of different spectral pretreatment methods show that the first-order pretreatment results are more ideal, and the average relative error of each nutrient can be reduced by more than 0.1%. The samples were divided into two intervals according to the concentration of Vc, and the average relative error was reduced by more than 1% compared with the model of the whole concentration interval. Using PLSR and KNN-KSR methods to predict nutrients in different concentrations of milk powder rapidly based on near-infrared spectral information, the prediction effect of KNN-KSR is better. In order to obtain accurate prediction results, it is suggested to use the same level of concentration samples to establish the near-infrared quantitative model of nutritional components of milk powder and to apply it in concentration range.