摘要
奶粉市场是食品掺假行为频发领域,其中婴幼儿配方奶粉价格高,其质量是消费者、生产企业和执法部门关注的重点.近红外高光谱成像(Near infrared-hyperspectral imaging,NIR-HSI)技术结合化学计量学和机器学习算法可以检测奶粉中单一掺假物含量.基于NIR-HSI技术研究了不同品牌婴幼儿奶粉中多掺假物(三聚氰胺、香兰素和淀粉)的定量预测.对基于像素点预处理后的高光谱图像划分感兴趣区域(Region of interest,ROI),提取ROI平均光谱.基于经典的过滤式特征选择算法拉普拉斯分数(Laplacian score)(无监督)和ReliefF(有监督)挑选建模关键变量,建立偏最小二乘回归模型(Partial least squares,PLS).开发包含自定义选择层的一维卷积神经网络模型(One-dimensional convolutional neural networks,1DCNN).自定义层根据权重系数绝对值,可确定重要波长变量.Laplacian score-PLS模型对预测集中奶粉、三聚氰胺、香兰素和淀粉质量分数预测结果均方根误差分别为0.1110％、0.057 0％、0.034 9％和0.348 1％.ReliefF-PLS模型对预测集中奶粉、三聚氰胺、香兰素和淀粉预测结果均方根误差分别为0.199 8％、0.054 0％、0.045 5％和0.182 3％.1DCNN模型对预测集中奶粉、三聚氰胺、香兰素和淀粉质量分数预测结果均方根误差分别为0.856 1％、0.091 1％、0.064 4％和0.294 2％.对Laplacian score、ReliefF和自定义选择层挑选出的前15个重要波长进行对比分析,不同特征选择方法挑选的特征波长子集有所区别,但都选择1 210、l 474、1 524、1 680 nm等附近波长.基于ReliefF-PLS模型的可视化结果表明了其良好的预测能力.
Abstract
Milk powder is the hardest hit area for food adulteration.Among them,infant formula milk powder is expensive and important,with quality being the focus of consumers,manufacturers,and law enforcement agencies.Near infrared-hyperspectral imaging(NIR-HSI)technology combined with chemometrics and machine learning algorithms can detect the content of single adulterant in milk powder.The quantitative prediction of multiple adulterants(melamine,vanillin and starch)in different brands of infant milk powder was studied based on NIR-HSI technology.The hyperspectral images after pixel wise pretreatment were divided into regions of interest(ROI),and the ROI average spectra were extracted.The key variables for modeling were selected based on the classic filtering feature selection algorithms,i.e.Laplacian score(unsupervised)and ReliefF(supervised).Partial least squares(PLS)regression was adopted to establish prediction models.A one-dimensional convolutional neural network(1DCNN)model with a self-defined selection layer was developed.The self-defined layer determined the important wavelength variables according to the multiplicative weight parameters learned after modeling.The root mean square errors of prediction set of Laplacian score-PLS models to predict milk powder,melamine,vanillin and starch were 0.111 0％,0.057 0％,0.034 9％and 0.348 1％,respectively.The root mean square errors of prediction set of ReliefF-PLS models to predict milk powder,melamine,vanillin and starch were 0.199 8％,0.054 0％,0.045 5％and 0.182 3％,respectively.The root mean square errors of prediction set of 1DCNN models to predict milk powder,melamine,vanillin and starch were 0.856 1％,0.091 1％,0.064 4％and 0.294 2％,respectively.The first 15 important wavelengths selected by Laplacian score,ReliefF and self-defined selection layer were compared and analyzed.The characteristic wavelength subsets selected by different feature selection methods were different,but the wavelengths near 1 210 nm,1 474 nm,1 524 nm and 1 680 nm were selected in more than one method.The visualization results based on the ReliefF-PLS model demonstrated good predictive ability.
基金项目
国家自然科学基金项目(32102087)
河北省省级科技计划项目(21344801 D)
河北省专业学位研究生教学案例建设项目(KGJSZ2022005)
国家科技期刊平台
NSTL
万方数据