血清胆红素的近红外光谱无创检测

血清胆红素的无创检测在疾病的预防、早期诊断与后期治疗等阶段都具有极其重要的作用.本研究提出了一种基于近红外光谱技术的无创血清胆红素新方法.通过采集舌尖的近红外反射光谱并运用偏最小二乘法对采集到的光谱数据进行建模,从而实现对血清胆红素的快速无创检测.将采集到的全部57例样本按照4∶1的比例分训练集和预测集,分别建立总胆红素(TBIL)、直接胆红素(DBIL)和间接胆红素(IBIL)的偏最小二乘回归模型.3个模型的相关系数分别为0.9922,0.9947和0.9486,预测均方根误差(RMSEP)分别为6.13,4.61和4.05 μmol/L.结果表明:舌尖近红外反射光谱结合偏最小二乘法可用于血清胆红素含量的无创检测,并且为其它血液成分的无创检测开辟了新路径.     

基于BP-神经网络的航空煤油总酸值近红外光谱快速检测

针对航空煤油中总酸值量较小,在近红外定量分析时有用信息易被干扰的问题,采用误差反向传播神经网络(BP-ANN)建立航空煤油总酸值近红外光谱分析模型.根据模型校正集预测偏差最小原则,确定了隐含层神经元个数、学习效率等参数.用建立的网络模型预测了验证集样品总酸值,预测的相关系数R2为0.9778,预测标准偏差(RMSEP)...     

基于平行因子分析法提高近红外无创血糖校正模型稳健性的研究

采用近红外光谱进行无创血糖检测时, 样品背景变动造成的预测集样本与校正集样本量测体系不一致的问题是导致预测精度低的原因之一. 提出一种将母体背景作为变量引入回归建模中, 结合各个母体背景下的样本光谱信息构建三维光谱矩阵以提高校正模型稳健性的分析方法. 将平行因子分析(PARAFAC)与多元线性回归(MLR)相结合, 对人体三层皮肤模型的蒙特卡罗模拟实验和葡萄糖水溶液及其混合物的离体实验进行了验证. 实验结果表明, 与传统的单一母体背景所建立的偏最小二乘模型相比, 将母体背景作为建模元素采用PARAFAC-MLR法所建立的校正模型具有更好的预测能力和稳健性.     

基于可见/近红外透射光谱的番茄红素含量无损检测方法研究

针对番茄内外部结构特征,搭建了可见/近红外透射检测系统,利用完整番茄透射光谱信息,对番茄红素含量进行无损伤快速检测研究。采集的原始光谱曲线经去趋势(DT)、标准正态变量变换(SNV)、多元散射校正(MSC)、归一化(NOR)、一阶导数(FD)预处理后分别用偏最小二乘(PLS)进行建模分析。其中SNV预处理后的模型效果最好,校正集和验证集相关系数分别为0.9771和0.9504,校正集和验证集均方根误差为0.9711和1.0496 mg/kg。为进一步提高模型的精度和稳定性,采用无信息变量消除法(UVE)、连续投影算法(SPA)、竞争性自适应重加权算法(CARS)3种方法单独或联合处理(UVE-SPA,UVE-CARS),对全光谱进行变量优选。经UVE-CARS处理后番茄红素预测模型效果最好,其校正集和验证集相关系数分别提高至0.9830和0.9741,均方根误差分别降低至0.6919和0.7680 mg/kg。最后,选用25个番茄样品对所建立模型进行了外部验证,UVE-CARS-PLS模型的预测集相关系数为0.9812,预测集均方根误差为0.7071 mg/kg,平均相对误差为4.3%。而作为比较的PLS模型的预测集相关系数为0.951,均方根误差为1.0610 mg/kg,平均相对误差6.0%,相比于全光谱PLS模型,UVE-CARS可以很大程度地简化模型,提高模型精度,降低检测的误差限。结果表明,基于自行搭建的番茄可见/近红外透射检测系统结合光谱处理方法,可以实现对生鲜番茄中番茄红素含量的快速、无损检测,为番茄红素定量检测提供了新方法。     

废旧涤/棉混纺织物近红外定量分析模型的建立及预测

将中红外光谱筛选出的598个纯涤、纯棉及涤/棉混纺样本采用GB/T 2910.11-2009法测定其涤、棉准确含量,其中校正集样本252个,验证集样本346个。使用便携式近红外光谱仪获取样本的原始近红外光谱(NIRS)。校正集样本依据回归系数的分布趋势和范围选取最佳建模谱区,并采用差分一阶导、S-G平滑和均值中心化相结合的方法对原始光谱进行预处理,利用偏最小二乘法(PLS)建立涤/棉混纺织物中涤含量的近红外(NIR)定量分析模型。同时分析了样本颜色对NIRS的影响,探讨了斜线光谱样本、奇异样本和不同组织结构织物对模型预测效果的影响。结果表明:利用PLS法建立的涤/棉混纺织物定量分析模型最优组合包含1个光谱区间和9个主成分因子,校正集相关系数(RC)为0.998,标准偏差(SEC)为0.908。为验证所建模型的有效性和实用性,对346个未参与建模的涤棉样本进行了预测,并将预测结果与国标法测定值进行方差分析,两种方法结果无显著差异,预测正确率达97%以上。模型的建立为废旧涤/棉混纺织物快速、无损分拣提供了基础数据库。     

近红外光谱法快速检测何首乌炮制过程中多糖含量的研究

该研究基于近红外光谱技术建立了何首乌在蒸制过程中多糖含量变化的快速定量模型。采用蒽酮-浓硫酸法测定了不同蒸制时间何首乌的多糖含量并结合采集的近红外光谱数据建立模型,以校正集相关系数(R²_c)、预测集相关系数(R²_p)、交叉验证集均方根误差(RMSECV)和预测集均方根误差(RMSEP)作为评价指标,考察了不同预处理方法和变量筛选方法的效果。结果显示,随着蒸制时间的延长,何首乌多糖含量先上升后下降并趋于平稳,采用平滑+变量标准化+随机蛙跳变量筛选建立的偏最小二乘法回归模型的R²_c=0.96,RMSECV=0.74,R²_p=0.96,RMSEP=0.28;外部预测相对偏差小于3.0%。所建立的多糖含量定量模型质量较高,预测能力较强,可为探索何首乌炮制工艺标准化及质量评价提供参考。     

紫外-可见-短波近红外漫反射光谱技术测定白芷中二氧化硫残留量的研究

建立紫外-可见-短波近红外漫反射光谱结合化学计量学测定白芷中二氧化硫残留量的方法。利用紫外-可见-短波近红外漫反射光谱技术并结合化学计量学建模预测二氧化硫残留量。偏最小二乘回归法(PLSR)建模优于支持向量回归法(SVR); Random Frog波段选择结合Auto-scaling预处理后PLS建模后预测效果最佳,校正集R~2为0. 99,交叉验证集R~2为0. 94,预测集R~2为0. 96。紫外-可见-短波近红外漫反射光谱结合化学计量学可以实现二氧化硫残留量的快速检测,为中药饮片的质量评价及监管提供一种技术手段。     

近红外光谱法测定30%TBP/正十二烷中硝酸含量

核燃料后处理工艺控制分析中,有机相中硝酸含量是一项重要的控制参数。通过研究TBP/正十二烷介质中硝酸的近红外光谱,将有机相样品的傅立叶变换近红外光谱与偏最小二乘回归法相结合,建立了含铀后处理有机相样品中硝酸浓度的测量方法。建立的定量校正模型的最佳校正标准偏差(RMSEC)、预测标准偏差(RMSEP)以及相关系数(r)分别为0.011,0.014,0.999。方法检出限为0.05 mol/L,测量结果的相对标准偏差不大于4%(n=6)。采用近红外分析法与滴定法对模拟样品进行测量,对测量结果进行t检验,结果表明两种方法的测定结果无显著性差异。所建方法无需样品预处理,可直接测量,分析速度快,结果准确,具有一定的实用性。     

基于交互式自模型混合物分析的近红外光谱波长变量优选方法

为了提高近红外光谱定量分析的预测精度和建模效率,提出了一种基于交互式自模型的混合物分析的波长优选方法,根据光谱各波长变量的纯度值和标准差值,选择含有用信息的波长变量,并引入相关权函数解决变量间共线性问题.通过依次迭代选择的变量建立定量校正模型,由交互验证均方根预测误差(RMSECV)确定最佳波长变量个数.应用该波长变量优选方法对具有不同葡萄糖含量的两组(四成分葡萄糖水溶液实验和人体血浆实验)近红外光谱数据进行分析,两组数据中分别只选择了全部变量的0.3%建立定量校正模型,其验证集葡萄糖浓度的均方根预测误差(RMSEP)分别减少为669和15 mg/L.与全谱范围及优选波段建立的定量校正模型比较,本方法能够通过波长变量优选最小化冗余信息、提高预测精度及建模效率.     

近红外光谱结合膜富集技术测定饮料中微量邻苯二甲酸二异辛酯的含量

提出了一种基于在线膜富集的近红外漫反射光谱技术,对饮料中的微量塑化剂邻苯二甲酸二异辛酯(DEHP)进行快速检测。采用聚醚砜膜对饮料中的DEHP进行富集,将富集DEHP的膜直接进行近红外漫反射检测。参考DEHP的透射近红外光谱,对波数进行选择,以4 420～4 060、4 700～4 540、6 040～5 600cm-1作为建模的波数区间。通过比较原始光谱、多元散射校正、一阶求导、二阶求导及其组合,考察了光谱预处理方法对模型的影响,用去一交互验证法建立了偏最小二乘(PLS)模型,并用所建立的校正模型对校正集样品进行了预测。结果表明,在选定的波数区间,当用一阶求导对校正集光谱进行预处理时,所建立的模型对校正集的预测效果最佳,在隐变量数为7时,对校正集所有样品的校正均方根误差(RMSEC)为0.188 7mg/L。用此模型对预测集样品进行预测时,DEHP的质量浓度在0.5～5.0 mg/L范围内,预测均方根误差(RMSEP)为0.232 4 mg/L,平均相对预测误差为6.29%。     

Building the quality into pellet manufacturing environment--feasibility study and validation of an in-line quantitative near infrared (NIR) method

The present study focuses on the implementation of an in-line quantitative near infrared (NIR) spectroscopic method for determining the active content of pharmaceutical pellets. The first aim was to non-invasively interface a dispersive NIR spectrometer with four realistic particle streams existing in the pellets manufacturing environment. Regardless of the particle stream characteristics investigated, NIR together with Principal Component Analysis (PCA) was able to classify the samples according to their active content. Further, one of these particle stream interfaces was non-invasively investigated with a FT-NIR spectrometer. A predictive model based on Partial Least Squares (PLS) regression was able to determine the active content of pharmaceutical pellets. The NIR method was finally validated with an external validation set for an API concentration range from 80 to 120% of the targeted active content. The prediction error of 0.9% (root mean standard error of prediction, RMSEP) was low, indicating the accuracy of the NIR method. The accuracy profile on the validation results, an innovative approach based on tolerance intervals, demonstrated the actual and future performance of the in-line NIR method. Accordingly, the present approach paves the way for real-time release-based quality system.     

基于近红外光谱的食用植物油中反式脂肪酸含量快速定量检测及模型优化研究

利用近红外光谱技术对食用植物油中反式脂肪酸(Trans fatty acids,TFA)含量进行快速定量检测,并通过波段选择、预处理方法、变量筛选及建模方法对TFA含量预测模型进行优化.采用AntarisⅡ傅里叶变换近红外光谱仪在4000~10000 cm-1光谱范围采集98个食用植物油样本的近红外透射光谱,然后采用气相色谱法测定TFA的真实含量.首先,对样本原始光谱进行波段、预处理方法优选;在此基础上,采用竞争自适应重加权法(Competitive adaptive reweighted sampling,CARS)筛选TFA相关的重要变量,最后应用主成分回归、偏最小二乘和最小二乘支持向量机方法分别建立食用植物油中TFA含量的预测模型.研究结果表明,近红外光谱技术检测食用植物油中的TFA含量是可行的,优化后的最佳预测模型的校正集和预测集R2分别为0.992和0.989,RMSEC和RMSEP分别为0.071%和0.075%.最佳预测模型所用的变量仅26个,占全波段变量的0.854%.此外,与全波段偏最小二乘预测模型相比,其预测集R2由0.904上升为0.989,RMSEP由0.230%下降为0.075%.由此表明,模型优化非常必要,CARS能有效筛选TFA相关的重要变量,极大减少建模变量数,从而简化预测模型,并较大提高预测模型的精度和稳定性.     

Rapid determination of pH in solid-state fermentation of wheat straw by FT-NIR spectroscopy and efficient wavelengths selection

In the work discussed in this paper we investigated the feasibility of determination of the pH of a fermented substrate in solid-state fermentation (SSF) of wheat straw. Fourier-transform near-infrared (FT-NIR) spectroscopy was combined with an appropriate multivariate method of analysis. A genetic algorithm and synergy interval partial least-squares (GA-siPLS) were used to select the efficient spectral subintervals and wavelengths by k-fold cross-validation during development of the model. The performance of the final model was evaluated by use of the root mean square error of cross-validation (RMSECV) and correlation coefficient (R (c)) for the calibration set, and verified by use of the root mean square error of prediction (RMSEP) and correlation coefficient (R (p)) for the validation set. The experimental results showed that the optimum GA-siPLS model was achieved by use of seven PLS factors, when four spectral subintervals were selected by siPLS and then 45 wavelength variables were chosen by use of the GA. The predicted precision of the best model obtained was: RMSECV = 0.0583, R (c) = 0.9878, RMSEP = 0.0779, and R (p) = 0.9779. Finally, the superior performance of the GA-siPLS model was demonstrated by comparison with four other PLS models. The overall results indicated that FT-NIR spectroscopy can be successfully used for measurement of pH in solid-state fermentation, and use of the GA-siPLS algorithm is the best means of calibration of the model.     

基于粒子群算法的最小二乘支持向量机在红花提取液近红外定量分析中的应用

提出一种基于粒子群算法的最小二乘支持向量机(PSO-LS-SVM)方法,用于建立红花提取过程关键质控指标的定量分析模型.近红外光谱数据经波段选择、预处理和主成分分析(降维)后,利用粒子群优化(PSO)算法对最小二乘支持向量机算法中的参数进行优化,然后使用最优参数建立固含量和羟基红花黄色素A(HSYA)浓度的定量校正模型.将校正结果与偏最小二乘法回归(PLSR)和BP神经网络(BP-ANN)比较,并将所建的3个模型用于红花提取过程未知样本的预测.结果表明,BP-ANN校正结果优于PSO-LS-SVM和PLSR,但是对验证集和未知样品集的预测能力较差,而PSO-LS-SVM和PLSR模型的校正、验证结果相近,相关系数均大于0.987,RMSEC和RMSEP值相近且小于0.074,RPD值均大于6.26,RSEP均小于5.70％.对于未知样品集,pSO-LS-SVM模型的RPD值大于8.06,RMSEP和RSEP值分别小于0.07％和5.84％,较BP-ANN和PLSR模型更低.本研究所建立的PSO-LS-SVM模型表现出较好的模型稳定性和预测精度,具有一定的实践意义和应用价值,可推广用于红花提取过程的近红外光谱定量分析.     

Comparative Study of Variable Selection Using Genetic Algorithm with Various Types of Chromosomes

In this study,different methods of variable selection using the multilinear step-wise regression(MLR) and support vector regression(SVR) have been compared when the performance of genetic algorithms(GAs) using various types of chromosomes is used.The first method is a GA with binary chromosome(GA-BC) and the other is a GA with a fixed-length character chromosome(GA-FCC).The overall prediction accuracy for the training set by means of 7-fold cross-validation was tested.All the regression models were evaluated by the test set.The poor prediction for the test set illustrates that the forward stepwise regression(FSR) model is easier to overfit for the training set.The results using SVR methods showed that the over-fitting could be overcome.Further,the over-fitting would be easier for the GA-BC-SVR method because too many variables fleetly induced into the model.The final optimal model was obtained with good predictive ability(R2 = 0.885,S = 0.469,Rcv2 = 0.700,Scv = 0.757,Rex2 = 0.692,Sex = 0.675) using GA-FCC-SVR method.Our investigation indicates the variable selection method using GA-FCC is the most appropriate for MLR and SVR methods.     

Support vector machines in water quality management

Support vector classification (SVC) and regression (SVR) models were constructed and applied to the surface water quality data to optimize the monitoring program. The data set comprised of 1500 water samples representing 10 different sites monitored for 15 years. The objectives of the study were to classify the sampling sites (spatial) and months (temporal) to group the similar ones in terms of water quality with a view to reduce their number; and to develop a suitable SVR model for predicting the biochemical oxygen demand (BOD) of water using a set of variables. The spatial and temporal SVC models rendered grouping of 10 monitoring sites and 12 sampling months into the clusters of 3 each with misclassification rates of 12.39% and 17.61% in training, 17.70% and 26.38% in validation, and 14.86% and 31.41% in test sets, respectively. The SVR model predicted water BOD values in training, validation, and test sets with reasonably high correlation (0.952, 0.909, and 0.907) with the measured values, and low root mean squared errors of 1.53, 1.44, and 1.32, respectively. The values of the performance criteria parameters suggested for the adequacy of the constructed models and their good predictive capabilities. The SVC model achieved a data reduction of 92.5% for redesigning the future monitoring program and the SVR model provided a tool for the prediction of the water BOD using set of a few measurable variables. The performance of the nonlinear models (SVM, KDA, KPLS) was comparable and these performed relatively better than the corresponding linear methods (DA, PLS) of classification and regression modeling.     

应用遗传算法和PLS的近红外光谱预测玉米中淀粉含量的研究

以普通玉米籽粒为试验材料,在应用遗传算法结合偏最小二乘回归法对近红外光谱数据进行特征波长选择的基础上,应用偏最小二乘回归法建立了特征波长测定玉米籽粒中淀粉含量的校正模型．试验结果表明,基于11个特征波长所建立的校正模型,其校正误差（RMSEC）、交叉检验误差（RMSECV）和预测误差（RMSEP）分别为0.30％、0.35％和0.27％,校正数据集和独立的检验数据集的预测值与实际测定值之间的相关系数分别达到0.9279和0.9390,与全光谱数据所建立的预测模型相比,在预测精度上均有所改善,表明应用遗传算法和PLS进行光谱特征选择,能获得更简单和更好的模型,为玉米籽粒中淀粉含量的近红外测定和红外光谱数据的处理提供了新的方法与途径．     

近红外光谱测定溶液中甲醇的含量

测量了含微量甲醇(体积分数为0.04%～0.24%)的系列乙醇水溶液的近红外光谱,利用近红外光谱分析建立了预测甲醇含量的定量分析模型。比较了用外部检验法(Test Set-Validation)和交叉检验法(Cross-Validaton)建立的数学模型,研究了使用外部检验法时,校正集和检验集样品数的改变对模型预测结果的影响。结果发现,当校正集样品数为15检验集样品数为6(总样品数为21)时,使用外部检验法建立的数学模型预测结果较好,其校正集的均方根误差和检验集的预测均方根误差(分别为RMSEE和RMSEP)均较小(分别为0.0115和0.0105),而且很接近。结果表明,近红外光谱方法简单,准确而且实用。     

多模型共识偏最小二乘法用于近红外光谱定量分析

将多模型共识偏最小二乘法用于近红外光谱定量分析。利用随机抽取的训练子集建立一系列偏最小二乘模型,选取其中性能较好的部分模型作为成员模型,用这些成员模型来预测未知样品。将该方法用于一组生物样本的近红外光谱与样品中人血清白蛋白、γ-球蛋白以及葡萄糖含量之间的建模研究,并与单模型偏最小二乘法了进行比较。结果 PLS对独立测试集中三种组分进行50次重复预测的平均RMSEP分别为0.1066,0.0853和0.1338,RMSEP的标准偏差分别为0.0174,0.0144和0.0416;而本方法重复预测的平均RMSEP分别为0.0715,0.0750和0.0781,RMSEP的标准偏差分别为0.0033,0.2729×10-4和0.0025。