近红外光谱技术快速检测小龙虾新鲜度

应用近红外光谱技术实现对小龙虾新鲜度的快速检测。利用化学计量学方法,通过对近红外品质分析仪采集的虾肉绞碎前后光谱（850～1 050 nm）调整不同预处理方法、偏最小二乘法和组合算法,建立一种基于总挥发性盐基氮（total volatile basic nitrogen,TVB-N）含量的小龙虾新鲜度定量预测模型。结果表明：采用标准正态变量变换与一阶导数结合的预处理方法模型预测效果最好,且绞碎后的虾肉光谱比绞碎前建模效果更好;为满足实际应用需要,对绞碎前的虾肉TVB-N含量预测模型进行分析,其交叉验证误差为3.123,交叉验证相关系数为0.947,用此模型对预测集24 个样品进行预测,预测值与实测值的交叉验证相关系数为0.951 4,在TVB-N含量超过20 mg/100 g（不新鲜）的检测准确率为100%。近红外光谱技术可应用于快速检测小龙虾新鲜度,所建模型具有较好的预测能力。     

近红外光谱技术快速鉴别原料肉掺假的可行性研究

探讨利用近红外光谱技术结合Fisher两类判别法以及多层感知器(multilayer perceptron,MLP)神经网络快速无损鉴别原料肉是否掺假,并建立多种掺假肉的分类识别模型的可行性.首先近红外结合主成分与Fisher两类判别,建立原料肉与掺假肉的判别函数,以原料肉与注水肉两类样木的平均重心即两类样木的加权平均...     

近红外光谱技术在生鲜禽肉质量检测中应用的研究进展

近红外光谱（near-infrared spectroscopy,NIRS）技术作为一种快速、无损、绿色的检测技术正在被广泛用于禽肉品质研究,通过将肉样光谱信息和品质指标参考值相关联,构建高精度、高稳定性的数学模型预测未知肉品品质。相比传统检测方式,NIRS技术具有无需预处理、更加丰富的信息量、数据计算更快等优势,在肉制品检测应用方面潜力巨大。本文主要综述了NIRS技术在生鲜禽肉（鸡肉、鸭肉、鹅肉）物理属性、化学指标以及微生物腐败等方面检测的研究进展,归纳总结了NIRS技术结合不同化学计量学算法构建模型检测禽肉各品质指标的效果,同时提出了NIRS技术在检测禽肉方面存在的缺点及未来发展趋势,可为改善NIRS技术检测应用、研发便携式NIRS检测设备提供数据支撑和理论参考。     

基于近红外光谱法的桑蚕丝接枝率快速定量测定

为解决蚕丝经化学接枝增重处理后,接枝率难以直接测定以及现有的热分析法检测耗时长、不适用于批量化快速检测等问题,提出了采用近红外光谱技术对蚕丝接枝率进行快速测定的方法。应用近红外光谱法结合化学计量学软件,选择偏最小二乘法,从光谱预处理、最佳主因子数选择以及建模谱区选择3个方面优化建立甲基丙烯酰胺接枝蚕丝的接枝率预测模型,得到所建模型的内部预测准确率为91.03%。使用19个已知参比值但未参与建模的样本对模型的稳健性进行验证,对预测值和参比值进行配对t检验,在给定显著水平α为0.05条件下,模型预测结果与称重法测试结果不存在显著性差异。结果表明,近红外光谱技术可为蚕丝接枝率的定量测定提供一种快速有效的分析方法。     

基于近红外光谱法快速测定许氏平鲉脂肪和 水分含量的方法研究

目的建立近红外光谱法结合偏最小二乘法测定许氏平鲉鱼肉中的脂肪和水分含量,以期简便、快速地对许氏平鲉进行品质分析与评价。方法采用常规分析手段测定70个样品的脂肪和水分含量,同时采集其近红外光谱数据,结合偏最小二乘法(partial least square,PLS)建立许氏平鲉鱼肉中脂肪和水分的定量预测模型,并对比不同光谱预处理方法、光谱范围和因子数对定量预测模型的影响。结果光谱经Savitzky-Golay(S-G)和标准正态变量变换(standardized normal variate,SNV)预处理后,在5341.85~4007.36 cm~(-1)、6556.79~5345.71cm~(-1)和8651.10~7162.33 cm~(-1)光谱范围内,选取主因子数10,建立脂肪的校正模型性能最优;光谱经过SNV预处理后,在8886.38~4061.35cm~(-1)光谱范围内,分别选取主因子数为9时,建立的水分的校正模型性能最优。脂肪和水分含量相对最优PLS模型的校正集相关系数分别为0.9918和0.9912,校正标准偏差分别为0.2680和0.3300,交叉验证相关系数分别为0.9820和0.9810,交叉验证均方差分别为0.3980和0.4850,验证集相关系数分别为0.9804和0.9798,验证集均方差分别为0.3260和0.3070。结论本方法可较为准确地预测许氏平鲉鱼肉中的脂肪和水分含量,能够满足快速分析评价许氏平鲉品质的要求。     

Chemical and discriminant analysis of bovine meat by near infrared reflectance spectroscopy (NIRS)

Near infrared reflectance spectroscopy (NIRS) was evaluated as a tool to segregate different types of bovine meat and predict several chemical fractions on samples from two breeds, three muscles and six grading (Chilean system) categories. Samples previously minced, frozen and thawed, were scanned (400–2500 nm) and then analyzed for dry matter, crude protein, ether extract, total ash and collagen content, after freeze drying. Discriminant analysis using a partial least squares regression technique and cross validation, correctly identified breed and muscle type for most samples, but carcass grades, with the exception of samples from calves, were not successfully predicted. Best calibrations for chemical composition tested by cross-validation, showed R² and standard errors of cross validation of 0.77 and 0.58% (dry matter), 0.82 and 0.48% (crude protein), 0.82 and 0.44% (ether extract). Calibrations for total ash showed a poor, and for collagen, a very poor prediction ability.     

近红外光谱法快速测定榨菜中亚硝酸盐含量

为研究利用傅立叶近红外光谱分析仪(NIRS)快速测定市售榨菜中亚硝酸盐的含量,先取榨菜样品按GB5009.33-2016测定其亚硝酸盐含量,再向榨菜样品中添加亚硝酸钠,制成亚硝酸钠浓度范围为0.122~39.0875 mg/kg,浓度梯度为0.66 mg/kg的60个样本校正集;与10个样本预测集采集对应的傅立叶近红外光谱曲线,将光谱信息与实际测量值相关联,利用TQ analyst建模软件进行计算分析。结果表明:建模最优预处理方法为一阶微分(1D)与Savitzky-Golay filter滤波平滑的组合预处理;比较分析偏最小二乘法(PLS)与主成分回归法(PCR)的亚硝酸盐样品建立的光谱模型,数据结果显示采用偏最小二乘法(PLS)的亚硝酸盐组分模型稳定性和预测能力更好;内部交叉验正均方差(RSMECV)、交叉验证决定系数(R_c)、外部预测均方根误差(RMSEP)、预测决定系数(R_P)相关系数(r)分别为0.0310、0.9925、0.0141、0.9720、0.9378。经F检验与t检验,与国标所测结果无显著性差异。NIRS检测快速,无损便捷,可用于市售榨菜中亚硝酸盐残留量的定量检测。     

鸡腿肌冻干粉蛋氨酸近红外光谱定量预测模型的建立与优化

本文从光谱预处理方法、建模特征光谱筛选、异常样本剔除、建模样本选择四个方面建立和优化鸡腿肌冻干粉蛋氨酸近红外定量预测模型,旨在进一步提高模型的预测精度和模型稳健性。以263个鸡腿肌冻干粉NIRS和蛋氨酸含量为研究对象,分别使用7种不同光谱预处理方法、4种特征光谱筛选方法、2种MCCV异常样本剔除方法,SPXY和常规选择2种建模样本选择方法,应用偏最二小乘法(PLS)、内部交互验证和外部验证建立和优化腿肌冻干粉蛋氨酸近红外定量预测模型。结果表明：在本研究中,最优鸡腿肌冻干粉蛋氨酸NIRS定量预测模型为在1000-2502nm谱段,使用原始光谱,在SNV+gapsegment(1#,15,7)光谱的基础上使用MCCV方法删除54个样本后,采用SPXY方法选取156个校正样本,39个外部验证样本所建模型,其为0.93、SE_CV为0.0609、为0.83、RPD_P为2.42。研究表明,模型预测值与化学检测值有很高的相关度,对腿肌冻干粉蛋氨酸NIRS模型预测精度和稳健性影响最大因素是异常样本剔除方法和建模样本选取方法。     

Classification of geographical origins and prediction of δ13C and δ15N values of lamb meat by near infrared reflectance spectroscopy

Near infrared spectroscopy (NIRS) combined with chemometric analysis was investigated for its potential to classify the geographical origin and predict δ(13)C and δ(15)N values of lamb meat samples (n=99) from three pastoral regions and two agricultural regions of China. Principal component analysis (PCA), discriminant partial least squares analysis (D-PLS), linear discriminant analysis (LDA) and partial least squares regression (PLSR) were used for data analysis. D-PLS and LDA correctly classified 100% of the both pastoral and agricultural region samples, and gave a total correct classification of 88.9% and 75% to the five individual region samples, respectively. The best PLSR calibration models for predicting δ(13)C and δ(15)N of lamb meat were obtained with the determination coefficient (R(2)) 0.76 and 0.87, respectively. These results show that NIRS combined with chemometrics can be used as a rapid and effective method to discriminate the geographical origin and estimate the δ(13)C and δ(15)N of lamb meat.     

Combination and comparison of multivariate analysis for the identification of orange varieties using visible and near infrared reflectance spectroscopy

A study of multivariate analysis for orange varieties was carried out, and the potential of visible and near infrared reflectance spectroscopy (Vis/NIRS) for its ability to nondestructively differentiate orange varieties was evaluated. A total of 320 orange samples (80 for each variety) were investigated for Vis/NIRS on 325–1075 nm using a field spectroradiometer. Multivariate classification methods including principal component analysis (PCA), back propagation neural network (BPNN) and partial least squares discriminant analysis (PLSDA) were adopted to classify oranges. Sixteen principal components from PCA were used as the input of BPNN model, and the identification accuracy of four orange varieties reached 100%. The prediction result of PLSDA, i.e., standard error of prediction (SEP) 0.24497, correlation coefficient (R) 0.97843, root mean square error of prediction (RMSEP) 0.24268, and identification accuracy 90% indicate that PLSDA is an alternative model for orange identification. With the comparison of these two models, it shows that BPNN combined with PCA obtained better classification effect than that of PLSDA. The overall results demonstrate that Vis/NIRS technology with multivariate analysis models is promising for the rapid and reliable determination for identification of orange varieties.     

植物油中掺煎炸动物油的近红外快速无损检测

利用近红外技术结合偏最小二乘法建立掺煎炸动物油植物油的定量模型。对100个植物油样本,通过近红外光谱仪扫描获得从10 000～4 000cm~(-1)的光谱信息。运用TQ-Analyst软件进行计算,选择全谱区,结合偏最小二乘法(PLS)算法,得到光谱最佳预处理方法为一阶导加Norris平滑。进行内部交叉验证,相关系数r为0.992 0,预测误差为3.11,且预测结果与真实值可通过t检验,说明模型是可行的。     

On-Site NIR Spectroscopy to Control the Shelf Life of Pork Meat

Control of meat shelf-life includes the time that it remains in the exhibitor of sale (such as the supermarket) until its rejection for the consumer, or withdrawal due to expiry date. Near infrared spectroscopy (NIRS) is one of the most promising techniques for large-scale meat quality control. This study investigated the potential of on-site NIRS portable instrumentation-based models to predict three microbiological parameters to establish if pork meat is acceptable or not for consumption (aerobic Mesophilous microorganisms, Enterobacteriaceae, and lactic acid bacteria) and pH to quality control food preservation and shelf-life extension on intact slices of pork meat packaged under two different modified atmospheres. NIR calibrations were developed by using an on-site Phazir instrument (Polychromix, Wilmington, MA, USA) in the range 1,600–2,400 nm. A total of 252 samples of pork meat slices were directly scanned twice in reflectance mode on trays, once before and another one after removing the film cover at 1, 3, 5, 7, 9, 12, and 15 days of storage. Results showed that spectra of meat acceptable or not for consumption have marked differences around 1,660 nm. NIRS quantitative prediction models showed r ² values between 0.19 and 0.65 for the microbiological parameters assayed. The developed NIRS methodology makes possible on-site prediction of microbiological status of pork meat with a standard error of cross-validation around 1 log cfu/g. Results have shown that modified atmosphere packaging has no influence on calibration statistics.     

基于NIRS的食用醋品牌溯源模型的建立与优化

本文主要探讨了近红外光谱(NIRS)结合模式识别技术应用于食用醋品牌溯源研究。采集了四个品牌(四川保宁香醋、山西东湖老陈醋、镇江恒顺香醋、镇江香醋)共160组食醋样品的近红外漫反射光谱,通过主成分分析(PCA)进行光谱变量压缩及剔除8个异常样本数据后,随机选取其中的114组样品组成训练集用于建立溯源模型,剩余38组样品用作测试集进行模型验证。比较了MSC、SD、SNV等几种不同光谱预处理方法以及它们的不同组合对溯源模型的影响,同时考察了PLS-DA与SIMCA两种建模方法对模型的影响。结果表明:选择MSC与SD相结合的方法对光谱数据进行预处理,并采用SIMCA建模方法所建立的醋品牌溯源模型对四大品牌醋的正确识别率分别可达100%、100%、91.7%、90%。由此说明采用近红外光谱技术结合模式识别技术可有效实现食用醋品牌溯源的目的。     

肉类电子鼻识别模型的建立

采用电子鼻对牦牛肉、牛肉和鸭肉样品进行检测,同时对经热处理的牦牛肉和牛肉样品进行分析.通过对所获得的数据进行主成分分析(principal component analysis,PCA)和判别因子分析(discriminant factor analysis,DFA),建立用于识别不同肉类的DFA模型,并对模型进行验证,DFA模型对于未知样品的识别率达到100%,能够有效的区分和识别牦牛肉、牛肉和鸭肉样品.电子鼻分析的结果显示,热处理对牦牛肉的挥发性物质影响较大,而对牛肉的影响较小.     

基于长短期记忆神经网络的重大活动中鲜(冻)肉制品铅含量风险预测

目的 针对未来时间内重大活动举办地鲜(冻)肉制品铅含量风险预测的问题,建立基于LSTM的时间序列预测模型,对当地鲜(冻)肉制品铅含量进行风险评估与预测预警。方法 通过收集2011-2020年国家市场监督管理总局日常食品监督管理抽检数据,筛选出北京的鲜(冻)肉制品的抽检数据,构建数据集并进行预处理,按照7:2:1的比例划分训练集、验证集和测试集,基于Tensorflow平台构建4层LSTM模型并进行训练,基于重大活动举办前10天的鲜(冻)制品铅含量数据,对未来1天的鲜(冻)制品铅含量风险进行预测。结果 实验表明,经过50轮模型迭代训练,训练集和测试集Loss指标收敛至0.084,经过5次训练后的模型评估参数RMSE为0.192,R2_score为0.916,模型误差较小、准确度较高。结论 基于LSTM的鲜(冻)肉制品铅含量风险预测模型整体性能较好,可应用于重大活动举办地的食品风险预测,并精准指导监督抽检。     

Determination of cashmere and wool contents in textile by near infrared spectroscopy北大核心

A quantitative analysis model was established to determine the contents of cashmere and wool in textile by near infrared spectroscopy (NIRS). 101 calibration samples were mixed by weighing and measured by NIRS, which were combined to establish the calibration model by partial least squares (PLS) method. The correlation coefficient between the predictive result and true value was 0.99, and the root mean standard error of prediction was 2. 8% . The results demonstrated that NIRS could be a rapid and nondestructive technique for quantitative analysis of cashmere and wool. However, extensive sample collections were suggested in the future work to set up a representative database and to establish a mathematical model with better stabilization, precision and dependability.     

低温贮藏过程中水牛肉品质变化研究

为探究低温贮藏过程中水牛肉品质变化规律,选取新鲜水牛肉为原料,采用4、-18、-60℃三个温度梯度贮藏肉样,测定不同贮藏期水牛肉的色泽、滴水损失、pH、双烯值、蒸煮损失、挥发性盐基氮(TVB-N值)、菌落总数和大肠菌群数的变化。结果发现,随着贮藏温度的升高,水牛肉的蒸煮损失、b*值、双烯值、挥发性盐基氮、菌落总数和大肠菌群数显著增加(P<0.05),但水牛肉的a*值、L*值显著降低(P<0.05);随着贮藏时间的延长,水牛肉的滴水损失、蒸煮损失、双烯值、挥发性盐基氮、菌落总数和大肠菌群数显著增大(P<0.05);这表明水牛肉品质劣变程度进一步加剧。三个贮藏温度条件下,4℃冷藏的水牛肉双烯值、TVB-N值、菌落总数比-18、-60℃冻藏的水牛肉增长速度快,其中贮藏第10 d的TVB-N值和菌落总数分别为20.08 mg/100 g和6.7lg (CFU/g),显著高于-18、-60℃冻藏水牛肉(P<0.05),且-18、-60℃冻藏水牛肉均在卫生标准(GB 2707-2016)范围内。综上,贮藏时间的延长与温度的升高均会导致水牛肉品质下降。4℃冷藏适合水牛肉的短期保藏,货架期为8 d;但-18、-60℃冻藏适合水牛肉的长期保藏,其中60℃冻藏更有利于水牛肉品质的稳定和卫生安全,同时有效延长水牛肉的货架期。     

Prediction of lamb meat fatty acid composition using near-infrared reflectance spectroscopy (NIRS)

The aim of this study was to assess the feasibility of near-infrared reflectance spectroscopy (NIRS) for predicting lamb meat fatty acid composition. We compared ground vs. intact non-ground meat samples to determine whether grinding and homogenisation of meat samples improved the performance of the predictions. We used 76 male lambs, of which 32 were pasture-fed and 44 stall-fed with concentrate and hay. The reflectance spectrum of Longissimus lumborum muscle was measured at wavelengths between 400 and 2500 nm. Predictions were better with ground than with intact muscle samples. NIRS accurately predicts several individual fatty acids (FA) (16:0, 18:0, 16:1 Δ9 cis, 17:1 Δ9 cis, 18:1 Δ9 cis, 18:1 Δ11 cis and 16:1 Δ9 trans) and several FA groups (total linear saturated FA, total branched saturated FA, total saturated FA, total cis monounsaturated FA (MUFA), total trans MUFA, total MUFA and total polyunsaturated PUFA). These results show the potential of NIRS as a rapid, and convenient tool to predict the major FA in lamb meat.     

基于近红外光谱的大豆水分和粗脂肪含量的快速检测

基于近红外光谱技术快速检测大豆中水分和粗脂肪含量。方法 首先采集350-2500 nm光谱范围的大豆近红外光谱,采用光谱-理化值共生距离（SPXY）算法将大豆样本划分为校正集样本与测试集样本,然后对原始光谱分别采用多元散射校正(MSC)、标准正态变量交换(SNV)、归一化（Nor）等9种方法进行预处理,最后使用偏最小二乘回归(PLSR)分析方法建立模型对样本进行定量分析。结果 原始光谱经过多元散射校正后建立的偏最小二乘回归模型对水分的预测精度最高,其校正集和测试集的相关系数分别为0.8964和0.9055 , 均方根误差分别为0.4211和0.5933;原始光谱经过归一化处理后建立的偏最小二乘回归模型对粗脂肪的预测精度最高,其校正集和测试集的相关系数分别为0.9084和0.9295 , 均方根误差分别为0.6897和0.6462。结论 近红外光谱（NIRS）结合预处理及偏最小二乘回归法,可以快速、准确的检测大豆水分和粗脂肪含量。     

Prediction of black tea fermentation quality indices using NIRS and nonlinear tools

Catechin content, the ratio of tea polyphenols and free amino acids (TP/FAA), as well as the ratio of theaflavins and thearubigins (TFs/TRs) are important biochemical indicators to evaluate fermentation quality. To achieve rapid determination of such biochemical indicators, synergy interval partial least square and extreme learning machine combined with an adaptive boosting algorithm, Si-ELM-AdaBoost algorithm, were used to establish quantitative analysis models between near infrared spectroscopy (NIRS) and catechin content and between TFs/TRs and TP/FAA, respectively. The results showed that prediction performance of the Si-ELM-AdaBoost mixed algorithm is superior than that of other models. The prediction results with root-mean-square error of prediction ranged from 0.006 to 0.563, the ratio performance deviation values exceeded 2.5, and predictive correlation coefficient values exceeded 0.9 in the prediction model of each biochemical indicator. NIRS combined with Si-ELM-AdaBoost mixed algorithm could be utilized for online monitoring of black tea fermentation. Meanwhile, the AdaBoost algorithm effectively improved the accuracy of the ELM model and could better approach the nonlinear continuous function.