倪晓锋,张寅升,周亚,赵亚菊,郭芳婕,王姗姗,王海燕

• 1:浙江食药质量安全工程研究院
• 2:浙江工商大学管理工程与电子商务学院
• 3:浙江大学生物系统工程与食品科学学院

摘要(Abstract)：

市场上销售的食用盐品种繁多，主要以海盐、井盐、岩盐和湖盐为原料来源，而不同来源食用盐的品质和价格相差甚远。该研究使用配备532 nm激光器的拉曼光谱仪，采集得到4种食用盐样品共80个原始拉曼光谱，采用主成分分析(PCA)和偏最小二乘(PLS)进行数据降维，Kennard-Stone(KS)算法将样本按3∶1划分为训练集与测试集后，结合K近邻(KNN)、支持向量机(SVM)和BP神经网络(BPNN)3种分类器，对4种不同食用盐的品种来源进行鉴别分析。结果表明，相较于原始光谱的分类模型对测试集的预测准确度在30%～50%和PCA的30%～40%之间，PLS-KNN、PLS-SVM和PLS-BPNN模型的预测准确度分别为90%、100%和100%。PLS降维后只需6个维度的信息即可保留原始变量信息解释性98%以上，并且PLS-SVM在建模速度快的同时保留了较高的分类精度和稳定性，为提高食用盐产品品质、改善评价标准和完善管理体系提供了技术支持。

关键词(KeyWords)： 拉曼光谱;主成分分析;偏最小二乘;K近邻;支持向量机;BP神经网络

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金资助项目(91746202,61806177);; 浙江省自然科学基金资助项目(LQ20C200004)

作者(Author): 倪晓锋,张寅升,周亚,赵亚菊,郭芳婕,王姗姗,王海燕

参考文献(References)：

• [1]PEGAH S V,JORG H,BERND H.Establishing a novel procedure to detect deviations from standard milk processing by using online Raman spectroscopy[J].Food Control,2022,131:108442.
• [2]JAKUB D,FATIH C A,ALEKSANDRA W,et al.Trends in biomedical analysis of red blood cells-Raman spectroscop against other spectroscopic,microscopic and classical techniques[J].TrAC Trends in Analytical Chemistry,2022,146:116481.
• [3]DU Y W,HAN D P,LIU S,et al.Raman spectroscopy-based adversarial network combined with SVM for detection of foodborne pathogenic bacteria[J].Talanta,2022,237:122901.
• [4]JOZEF R,VIT S,VERONIKA H Z,et al.Pattern recognition as a new strategy in high-resolution spectroscopy:application to methanol OH-stretch overtones[J].Physical Chemistry Chemical Physics:PCCP,2021,23(69):20193-20200.
• [5]MA D D,WANG L J,JIN Y B,et al.Application of UHPLC fingerprints combined with chemical pattern recognition analysis in the differentiation of six Rhodiola species[J].Molecules,2021,26(22):6855.
• [6]TIZIANO Z,MARK G K,FLORENCIO C B,et al.Instrumental odour monitoring system classification performance optimization by analysis of different pattern-recognition and feature extraction techniques[J].Sensors,2021,21(1):114.
• [7]CASTURA J C,RUTLEDGE D N,ROSS C F,et al.Discriminability and uncertainty in principal component analysis (PCA) of temporal check-all-that-apply (TCATA) data[J].Food Quality & Preference,2022,96:104370.
• [8]ALI A,MARGETTS B M,ZAINUDDIN A A.Exploration of the principal component analysis (PCA) approach in synthesizing the diet quality of the Malaysian population[J].Nutrients,2021,13(1):70.
• [9]LIU Y S,RAYENS W.PLS and dimension reduction for classification[J].Computational Statistics,2007,22(2):189-208.
• [10]LI T Y,WU Y Y,WU F,et al.Sleep pattern inference using IoT sonar monitoring and machine learning with Kennard-stone balance algorithm[J].Computers & Electrical Engineering,2021,93:107181.
• [11]PARK D J,SUPEKAR O D,GREENBERG A R,et al.Real-time monitoring of calcium sulfate scale removal from RO desalination membranes using Raman spectroscopy[J].Desalination,2021,497:114736.
• [12]赵忠光.基于拉曼光谱的硫酸盐定量方法研究[D].北京：华北电力大学，2020.
• [13]GARDNER D W,LI J Q,MORSHEDIFARD A,et al.Silicate bond characteristics in calcium-silicate-hydrates determined by high pressure Raman spectroscopy[J].Journal of Physical Chemistry C,2020,124(33):18335-18345.
• [14]FARSANI M H,DARBANI S M R,MOBASHERY A.Application of deep Raman spectroscopy to detect ammonium nitrate concealed in color fabrics[J].Vibrational Spectroscopy,2022,121:103405.
• [15]ZHU X Y,YING C Z,WANG J Y,et al.Ensemble of ML-KNN for classification algorithm recommendation[J].Knowledge-Based Systems,2021,221:106933.
• [16]SINLAE A A J,ALAMSYAH D,SUHERY L,et al.Classification of broadleaf weeds using a combination of K-nearest neighbor (KNN) and principal component analysis (PCA)[J].Sinkron,2022,7(1):93-100.
• [17]HE Y,ZHANG W,MA Y C,et al.The classification of rice blast resistant seed based on Raman spectroscopy and SVM[J].Molecules,2022,27:4091.
• [18]DING Y H,YAN Y L,LI J,et al.Classification of tea quality levels using near-infrared spectroscopy based on CLPSO-SVM[J].Foods,2022,11:1658.
• [19]ZHANG D H,LOU S.The application research of neural network and BP algorithm in stock price pattern classification and prediction[J].Future Generation Computer Systems,2021,115:872-879.
• [20]赵志磊，王雪妹，刘冬冬.基于BP-ANN和PLS的近红外光谱无损检测李果实品质的研究[J].光谱学与光谱分析，2022,42(9):2836-2842.