近红外光谱结合不同偏最小二乘法快速检测镇江香醋的浑浊度

利用近红外光谱(4000cm-1～10000cm-1)结合化学计量学方法快速检测了镇江香醋中的浑浊度。首先,用近红外光谱仪采集香醋样本的近红外光谱数据以及用离心法测定样本的浑浊度值;然后,采用间隔偏最小二乘法(iPLS)、反向区间偏最小二乘法(biPLS)、联合间隔偏最小二乘算法(siPLS)优选光谱特征区间;最后,采用全光谱(4000cm-1～10000cm-1)偏最小二乘法(PLS)对优选出来的区间建立香醋浑浊度近红外光谱模型。结果表明,采用siPLS将全光谱均匀划分30个子区间,选择4个子区间[4 10 18 27]联合时,建立的模型预测效果最佳,其RMSECV和RMSEP分别为0.173和0.208,校正集和预测集相关系数分别为0.9337和0.9004。因此,利用近红外光谱技术快速检测香醋中的浑浊度是可行的。     

近红外光谱结合区间偏最小二乘法应用于花生油酸价的测定

采用近红外光谱法结合不同区间偏最小二乘波长筛选法建立花生油酸价的定量分析模型。采用酸碱滴定法测定花生油样本的酸价同时采集近红外光谱数据;采用区间偏最小二乘法(iPLS)、向后区间偏最小二乘法(BiPLS)、移动窗口偏最小二乘法(mwPLS)优选光谱特征区间;采用偏最小二乘法(PLS)对优选出来的谱段建立酸价的定量模型。结果表明,采用mwPLS选择的谱段建立的模型预测效果最佳,RMSECV和RMSEP分别为0.247 76和0.131 5,校正相关系数和预测相关系数分别为0.993 2和0.996 9。因此,近红外光谱结合移动窗口偏最小二乘法可以快速准确测定花生油的酸价。     

茶皂素提取过程的原位实时监测技术

采用近红外光谱技术结合化学计量学方法,实时原位监测茶皂素提取过程中的茶皂素质量浓度、蛋白质质量浓度和多糖质量浓度的变化。研究结果表明,一阶求导(1^stDer)为最优的光谱预处理方法,3个指标的校正集决定系数(R_C)和预测集决定系数(RP)皆最高,交叉校验残差均方根误差(RMSECV)和预测残差均方根误差(RMSEP)最小。相较于偏最小二乘法(PLS)和区间偏最小二乘法(iPLS)算法对回归模型进行校正,以R_P和RMSEP为评价指标,联合区间偏最小二乘法(SiPLS)算法下所建模型稳健性最佳。茶皂素浓度模型中的R_P=0.9889,RMSEP=1.36;蛋白质浓度的R_P=0.9859,RMSEP=0.354;多糖浓度的_RP=0.9919,RMSEP=0.359。结论：近红外光谱技术联合Si-PLS模型可较好的实时监测茶皂素提取过程。     

基于siPLS的猕猴桃糖度近红外光谱检测

为了探寻一种快速无损检测猕猴桃糖度的方法,利用小波滤噪法对猕猴桃1000～2500nm 近红外光谱进行了预处理,并用偏最小二乘法(PLS)、区间偏最小二乘法(iPLS)和联合区间偏最小二乘法(siPLS)分别建立预测模型。结果表明,采用联合区间偏最小二乘法将光谱划分为16 个子区间,利用其中的第9、11、13 号3 个子区间联合建立的糖度模型效果最佳,其校正集相关系数和均方根误差分别为0.9414 和0.3788。预测集相关系数和均方根误差分别为0.9295 和0.3904,主因子数为7 个。研究表明,用小波滤噪和联合区间偏最小二乘法所建立的猕猴桃糖度模型不但减少建模运算时间,剔除噪声过大的谱区,而且预测能力和精度均有所提高。     

莲藕淀粉含量的近红外光谱无损检测方法

应用近红外光谱技术无损检测莲藕的淀粉含量。对光谱数据的3种预处理方法进行了比较分析,再采用偏最小二乘法(PLS)和联合区间偏最小二乘法(SiPLS)建立了莲藕淀粉含量的近红外光谱分析模型。研究结果表明,经多元散射校正、一阶导数和平滑等结合的预处理,采用联合区间偏最小二乘法(SiPLS)建立的模型最佳;其校正集的相关系数(Rc)和均方根误差(RMSEC)分别为0.960 0和0.741 6,预测集的相关系数(Rp)和均方根误差(RMSEP)为0.923 8和1.050 6,可以满足实际应用要求。结论:利用近红外光谱技术对莲藕淀粉含量进行无损检测切实可行。     

基于近红外光谱技术的海参品质快速检测

目的应用近红外光谱技术建立海参产地区分和胶原蛋白快速检测的方法。方法总计43个海参样品来自大连、福建、连云港、山东4个地区。首先采集样品的近红外光谱图,经过标准正态变量(standard normal variables,SNV)预处理,利用不同定性判别模型对海参产地进行区分。通过分光光度计法测定海参的胶原蛋白含量,利用偏最小二乘法(partial least squares,PLS)、区间偏最小二乘法(interval partial least squares,iPLS)、向后区间偏最小二乘法(backwards interval partial least squares,BiPLS)和联合区间偏最小二乘法(synergy interval partial least squares,Si PLS)建立了海参胶原蛋白含量的预测模型。结果产地区分模型中最小二乘支持向量机(least-squares support vector machine regression,LS-SVM)的识别率最高,校正集识别率为100%,预测集识别率为95.35%;海参胶原蛋白预测模型中BiPLS的预测效果较好,校正集相关系数Rc为0.9002,预测集相关系数Rp为0.8517。结论近红外光谱技术可实现对海参的产地区分和胶原蛋白的快速检测。     

蚁群-遗传算法优化近红外光谱检测花茶花青素 含量的研究

目的：花青素是鲜花花草茶中起主要保健作用的成分,本研究提出一种基于蚁群-遗传区间偏最小二乘法(ACO-GA-iPLS)的近红外谱区筛选方法,并将其用于花茶花青素含量的预测。方法：将蚁群算法(Ant Colony Optimization, ACO)和遗传算法(Genetic Algorithm, GA)相结合优选特征光谱区间,然后用区间偏最小二乘法算法(iPLS)建立光谱模型。首先对花茶近红外光谱进行预处理;然后用ACO-iPLS优选出特征子区间;最后对所选的特征子区间,用GA-iPLS进一步细化花青素的特征子区间,并建立花青素的预测模型。结果：优选出3个特征子区间(第1、9、10子区间),所建模型对应的交互验证均方根误差(RMSECV)和预测均方根误差(RMSEP)分别为0.1460mg/g和0.1840mg/g,校正集和预测集相关系数分别为0.9187和0.8856。结论： ACO-GA-iPLS可以有效选择近红外光谱特征波长,简化模型,提高模型精度。     

基于近红外光谱的大黄鱼新鲜度评价模型

目的 探索定量评价大黄鱼新鲜度的方法。方法 在整鱼背部采集近红外光谱, 将原始光谱预处理后分别与挥发性盐基氮(TVB-N)、菌落总数建立偏最小二乘(PLS)模型、区间偏最小二乘(iPLS)模型、向后区间偏最小二乘(biPLS)模型和联合区间偏最小二乘(siPLS)模型。结果 biPLS模型的精度最高、预测性能最佳。TVB-N的biPLS模型的校正集和预测集相关系数分别为0.8371和0.7652; 菌落总数的biPLS模型的校正集和预测集相关系数分别为0.878和0.7009。结论 大黄鱼的近红外光谱信息与其TVB-N、菌落总数间都存在较高的相关性, 所建模型可以快速、无损地定量评价大黄鱼的新鲜度。     

基于近红外光谱的猪肉蛋白质及脂肪含量检测

蛋白质及脂肪是猪肉的重要营养成分。随着人们对饮食健康的要求越来越高,对猪肉蛋白质及脂肪含量快速检测也成为必然。通过近红外光谱技术对猪肉进行光谱数据采集,将光谱数据分为校正集样本和预测集样本,然后,在MATLAB中利用多元散射校正(MSC)与均值中心化相结合的方法进行光谱预处理并采用联合区间偏最小二乘方法(SiPLS)获得猪肉蛋白质及脂肪含量与光谱数据特征之间的对应关系,从而定量分析猪肉蛋白质及脂肪含量。实验结果表明,建立的SiPLS检测脂肪及蛋白质含量预测模型的最优组合分别为划分为20个光谱区间并联合4个子区间和9个主成分因子,和划分19个光谱区间并联合4个子区间和10个主成分因子。其预测集的相关系数分别为0.9798、0.9788,交互验证均方根误差分别为0.228,0.241。研究结果表明,利用近红外光谱结合SiPLS算法可以快速准确检测猪肉蛋白质与脂肪含量。     

蚁群和遗传算法优化花茶花青素近红外光谱预测模型的比较

以建立花茶花青素含量的最优近红外光谱模型为目标,对比研究了蚁群算法(Ant ColonyOptimization,ACO)和遗传算法(Genetic Algorithm,GA)优化近红外光谱谱区的效果。ACO-i PLS将全光谱划分为12个子区间时,优选出第1、9、10共3个子区间,所建的校正集和预测集相关系数分别为0.901 3和0.864 2;交互验证均方根误差(RMSECV)和预测均方根误差(RMSEP)分别为0.160 0 mg/g和0.202 0 mg/g;GA-i PLS将全光谱划分为15个子区间时,优选出第1、5共2个子区间,所建模型的校正集和预测集相关系数分别为0.906 3和0.879 3,交互验证均方根误差(RMSECV)和预测均方根误差(RMSEP)分别为0.156 0 mg/g和0.206 0 mg/g。研究结果表明:ACO-i PLS和GA-i PLS均可以有效选择近红外光谱特征波长,其中GA-i PLS模型的精度更高。     

近红外光谱分析技术在植物蛋白饮料定量分析中的应用

利用近红外光谱分析技术对植物蛋白饮料中脂肪和可溶性固形物含量进行定量分析。采用向后间隔偏最小二乘法（BiPLS）、组合间隔偏最小二乘法（SiPLS）、遗传偏最小二乘法（GA-PLS）、竞争性自适应重加权法（CARS）优选波段,并结合偏最小二乘法（PLS）建立植物蛋白饮料中脂肪和可溶性固形物的定量分析模型。结果表明,4种方法对模型均有优化效果,可提高模型的稳定性和精准性,其中GA-BiPLS、GA-SiPLS优化效果最为明显,脂肪、可溶性固形物的决定系数R2分别达到了0.984、0.97和0.988、0.990,预测标准均方差（RMSEP）分别为0.026、0.030和0.170、0.155,相对分析误差（RPD）分别为8.077、7.000和9.112、10.000。表明近红外光谱技术作为一种快速、便捷的检测手段,适用于植物蛋白饮料品质的快速检测分析。     

Classification of monofloral honeys from different floral origins and geographical origins based on rheometer

A rheometer was used to classify commercial honeys. Five kinds of Yichun honeys from different floral origins and five kinds of Acacia honeys from different geographical origins were classified based on a rheometer by four pattern recognition techniques: Principal Component Analysis (PCA), Cluster Analysis (CA), Partial Least Squares (PLS), and Support Vector Machines (SVM). All the samples for different floral origins or different geographical origins were demarcated clearly by PCA, PLS. The samples from different floral origins could be classified by SVM, and the samples from different geographical origins also have a high correct classification rate (97.5%). The classification rates for different floral origins and geographical origins were 95% and 97.50% by CA, respectively. Three regression models: Principal Component Regression Analysis (PCR), Partial Least Squares Regression (PLSR), Support Vector Regression (SVR) were used for category forecast. The regression analysis showed that SVR with radial basis function kernel worked most effective.     

基于可见/近红外高光谱成像技术的鸡蛋新鲜度无损检测

以400~1 000nm高光谱系统获得鸡蛋样本的高光谱图像,利用蒙特卡洛法检测异常样本,采用不同预处理方法处理原始光谱;应用竞争性正自适应加权算法(Competitive Adaptive Reweighted Sampling,CARS)、遗传偏最小二乘法(Genetic Algorithms PLS,GAPLS)和间隔蛙跳法(Interval Random Frog,IRF)对预处理后光谱数据提取特征波长;分别建立基于全光谱和特征波长的偏最小二乘回归(Partial Least Squares Regression,PLSR)和最小二乘支持向量机(Least Squares Support Vector Machine,LS-SVM)鸡蛋新鲜度预测模型。结果表明:标准正态变量变换(Standardized Normal Variate,SNV)法为最优预处理方法;利用CARS、GAPLS和IRF分别选出8,35,74个特征波长;基于GAPLS提取的特征波长的LS-SVM模型最优,其校正相关系数(Rc)为0.899,预测相关系数(Rp)为0.832。表明基于高光谱成像技术的鸡蛋新鲜度无损检测是可行的。     

卷烟加料中1,2-丙二醇的快速测定方法

为快速评价卷烟加料均匀性,依据标记物1,2-丙二醇近红外光谱信息的特征,建立了基于1,2-丙二醇近红外预测模型的卷烟加料均匀性快速测定方法.研究表明:模型的决定系数R2 ＞0.95,对盲样预测相对误差基本小于5％;测定了3个牌号卷烟,其加料均匀性系数分别为87.9％、89.1％和90.9％,RSD＜ 3％.     

A Feasibility Study on the Evaluation of Quality Properties of Chinese Rice Wine Using Raman Spectroscopy

Rapid analysis of Chinese rice wine (CRW) is an important activity for quality assurance and control investigations. In recent years, due to its insensitivity to water and fewer overlapped bands, Raman spectroscopy (RS) may provide more useful qualitative and quantitative information on functional groups of various chemical compounds in CRWs than the conventional spectroscopic technique (e.g., infrared spectroscopy); there has been a growing interest in the application of RS in the qualitative and quantitative analysis in food industry. In this study, the applicability of RS hyphenated with chemometrics using different pretreated spectra was examined to develop rapid, low-cost, and non-destructive method for quantification of four enological parameters involved in CRW quality control. Partial least square (PLS) was used for building the calibration models for the four chemical parameters based on the full RS spectrum. The model was also optimized by using efficient wavelength selection algorithm, i.e., synergy interval partial least square (SiPLS) algorithm. In addition, soft independent modeling of class analogy (SIMCA) and linear discriminant analysis (LDA) were used as classification techniques to predict the brands (wineries) of CRW samples. The results demonstrated that compared with the PLS model using all wavelengths of RS spectra, the prediction precision of model based on the spectral variables selected by SiPLS was significantly improved with high values of the coefficient of determination (>0.90), residual predictive deviation (>3.0), and range error ratio (>10) for all of the four quality parameters. The SIMCA and LDA results, characterized by high percentages of correct classification (96.67 and 100.00 % as average value in prediction for SIMCA and LDA, respectively), showed that samples belonging to a particular brand could be correctly classified. The overall results indicated the suitability of RS combined with efficient variable selection algorithm to rapidly control the quality of CRW.     

基于矿物元素指纹分析中国地理标志猪肉的产地溯源

本研究初步探讨了矿物元素指纹分析对我国地理标志猪肉产地溯源的可行性。采用电感耦合等离子体质谱对四川巴山的青峪黑猪肉,山东莱芜黑猪肉和北京黑六猪肉三种不同地域来源的地理标志猪肉中33种矿物元素含量进行测定。通过元素含量筛选,排除猪肉样品中含量低于或接近检出限的元素,筛选出13个元素进行研究。结合主成分分析、多重比较分析和判别分析对数据进行统计分析。其中Na、Fe、Co、Cu、Zn、Se、Rb、Sr共8种元素在地域之间差异显著(p<0.05)。通过对比偏最小二乘法判别分析(PLS-DA)、基于正交信号校正的偏最小二乘判别分析(OPLS-DA)、支持向量机、朴素贝叶斯、决策树和神经网络六种分类模型,得出OPLS-DA和决策树分类模型较适合基于矿物元素指纹分析的特色猪肉产地溯源。矿物元素指纹分析在我国地理标志猪肉产地溯源领域具有长远的应用前景。     

Direct determination of polymerised triacylglycerides in deep-frying vegetable oil by near infrared spectroscopy using Partial Least Squares regression

A green method for the determination of polymerised triacylglyceride (PTG) in deep-frying vegetable oils of different botanic origin has been developed employing near infrared (NIR) spectroscopy and Partial Least Squares (PLS) regression. Four different types of oil were heated during several hours, with and without the addition of foodstuff. NIR transmission spectra were obtained directly from sample aliquots stored in glass vials, thus avoiding the consumption of solvents and minimising waste generation. Variables employed for building the PLS models were selected applying interval PLS (iPLS) as well as Uninformative Variable Elimination-PLS (UVE-PLS). A global PLS model using spectra of all four types of oils was compared to PLS models established for each oil type. Due to the small differences observed in the NIR spectra that can be related to the different botanic origin and results obtained from the PLS model comparison, the use of a global PLS model is recommended leading to prediction errors of 2.28% (w/w) for the determination of PTG in oils employed for frying different kinds of foods.