光谱预处理对棉涤混纺面料近红外定量模型的影响

以46个棉涤混纺面料样品为研究对象,采集样品的近红外漫反射光谱,光谱范围为12 000~4 000 cm-1,利用偏最小二乘法建立定量校正模型,并用交叉检验法对模型进行检验,以交叉验证均方差RMSECV和决定系数R2作为判断模型优劣的标准.对利用无光谱预处理、一阶导数法、二阶导数法、多元散射校正和矢量归一化五种不同预处理方法所建的模型进行了比较,发现对光谱进行矢量归一化预处理所建模型最优;此外还分析了建立纺织布料的近红外光谱定量分析模型时主要的误差来源及近红外光谱分析技术用于纺织面料定量分析的可行性.     

近红外光谱结合模式识别算法溯源识别卷烟纸油污

为实现对油渍污染卷烟的快速识别及溯源。采集卷烟生产过程中使用的6种润滑油污染在4种卷烟纸上样品的近红外光谱,结合未污染样品的近红外光谱,优选出原始样品原始光谱及二阶导数光谱中共有的差异性光谱波段范围,采用模式识别算法(principal component analysis-mahalanobis distance, PCA-MD)分别建立污染及未污染样品模式识别模型和不同卷烟上6种润滑油的模式识别模型。结果表明:1)基于污染及未污染样品原始近红外光谱及二阶求导光谱差异性分析,优化出的模式识别模型建模波数范围是:6 000～5 300 cm~(–1)和4 500～4 000 cm~(–1);2)建立的污染及未污染模式识别模型前3个主成分累计得分贡献率97.826%,模型分类效果明显,建模集及外部验证集样品的识别准确率均为100%;3)分别建立的6种油渍在4种卷烟纸上的溯源类模型,前3个主成分得分累计贡献率均大于96%,模型分类效果明显,建模集及外部验证集样品的识别准确率均为100%。所建立的基于近红外光谱分析方法结合模式算法(PCA-MD)可实现卷烟生产过程出现的"黄斑烟"中油渍烟的快速识别及污染油渍溯源。     

近红外透射光谱聚类分析快速鉴别食用油种类

以8种食用油纯油的43个样品为对象,研究了近红外透射光谱结合聚类分析法快速鉴别食用油种类的可行性.采集样品在12 500~4 000 cm-1范围内的傅立叶变换近红外透射光谱,利用光谱模式识别法中的聚类分析法对图谱进行定性分类鉴别.实验证明,光谱经二阶导数预处理后,最短距离法、最长距离法和方差平方和法均可准确无误地将食用油样品分为8类,判别模型对预测集样品的准确率达到100%.研究表明,近红外透射光谱结合聚类分析法可以为快速无损鉴别食用油种类提供一种准确可靠的方法.     

近红外光谱技术对滤棒中三醋酸甘油酯含量快速测定的研究

本文将卷烟滤棒的近红外光谱图与相应的三醋酸甘油酯含量进行关联,采用偏最小二乘法建立近红外光谱图与卷烟滤棒中三醋酸甘油酯含量间的数学校正模型,使用校正模型不需对样品进行前处理,并与CORESTA推荐方法测得的结果无显著性差异。该方法过程简单、快速,结果令人满意,适用于成型车间批量检测样品的需求。     

近红外漫反射光谱法快速检测火药吸收药混合液中黑索今组分含量

研究了采用近红外( NIR)漫反射光谱技术快速检测火药吸收药混合液中黑索今( RDX)组分含量的方法。将装有混合液样品的烧杯置于光谱仪主机光源窗口上方,直接采集样品光谱图。通过分析纯RDX和样本的近红外光谱,确定908~945 nm、1094~1253 nm和1577~1678 nm作为建模谱区。通过比较不同的光谱预处理办法的效果,选择标准正态变换( SNV )＋一阶导数＋谱线平滑对原始光谱进行预处理。采用偏最小二乘方法(PLS)对RDX组分建立了定量线性校正模型。结果表明：模型的交叉验证决定系数(R2cv)为0.9879,交叉验证均方根误差(RMSECV)为0.2420,预测均方根误差(RMSEP)为0.2127,预测结果的平均相对误差为0.5661％,25 s内完成样品分析。该方法可给改性双基火药的连续自动化生产提供技术支持。     

近红外光谱结合神经网络判别茶鲜叶产地

为快速、无损的判别鲜叶产地,维护恩施玉露的地理标志产品属性,采集恩施市芭蕉乡、白果乡和咸丰县茶鲜叶近红外光谱,经光谱预处理后,对校正集66个样品光谱数据进行主成分分析,然后建立BP神经网络预测模型,对验证集鲜叶样品的产地进行了预测,建立了8(输入节点)-4(隐含层节点)-1(输出节点)三层网络模型,验证集样品判别准确率为100%.近红外光谱技术结合神经网络能够快速、准确地判别茶鲜叶产地.     

近红外光谱技术检测烟叶中的氨基酸

研究应用傅里叶变换近红外光谱法快速测定烟叶中氨基酸含量的可行性,使用偏最小二乘法(PLS)为建模方法,选择3800~8000 cm-1谱段,采用二阶导数和Norris Derivative滤波法进行光谱预处理,建立了烟叶中氨基酸含量的近红外预测模型。采用留一(leave-one-out)交叉验证法进行建模,并以校正集样品的交叉验证相关系数(R)和均方差(RMSECV)为指标优化光谱预处理方法和模型参数,确定最佳预测模型。将近红外光谱技术与常规标准检测方法相比较,结果表明,近红外光谱技术可以较为准确的测定烟叶中氨基酸的含量。     

近红外快速检测烤烟等级质量与关键化学指标的关系

为客观、稳定的评价烤烟等级质量,该文以2017-2018年全国14个省市的烟叶样品为试验材料,以近红外光谱技术快速检测得到的主要化学成分为基础,以外观特征进行分类后的等级作为参照,探究糖氮比、烟碱、氮碱比等关键化学指标与烤烟外观等级质量的关系。结果表明:糖氮比、烟碱、氮碱比与等级质量存在非线性关系;并应用对勾、二次型等非线性分段拟合函数和加权求和的方法,综合糖氮比、烟碱、氮碱比建立基于化学成分的烤烟等级质量评价方法,并验证在烟碱值适中时,糖氮比值越高烤烟等级质量越好的规律。实验表明该文提出的方法模型具有较好稳定性。因此,应用近红外快速检测技术得到的主要关键化学指标可以实现烤烟等级质量客观、稳定的分析评价。     

近红外漫反射光谱结合神经网络检测小麦蛋白质含量

为了实现对小麦蛋白质含量的快速检测,提出了基于近红外光谱结合神经网络的小麦蛋白质检测方法.以160个小麦样品为对象,采集其近红外漫反射光谱,并以国标法分析小麦样品蛋白质含量,作为参考值.样品随机分成预测样品集和定标样品集,其光谱经标准归一化、去趋势等预处理后,采用BP神经网络和偏最小二乘法分别建立蛋白质含量定标模型.BP神经网络模型的预测相关系数和预测均方根误差分别为0.98和0.270 4%.而偏最小二乘法模型的预测相关性系数和预测均方根误差分别为0.98和0.303 8%.结果表明,两种方法建立的模型都具有较好的预测相关性和预测效果,其中BP神经网络模型优于偏最小二乘法模型.用非线性BP神经网络结合相应算法建立模型检测小麦蛋白质含量的定标模型可以提高检测准确性.     

包装方式对片烟醇化期间化学成分的影响

该文主要研究在真空包装与常规包装2种包装方式下,片烟醇化效果的差异性。选用同批样品经过2种不同方式包装后醇化3年,醇化结束后通过连续流动分析法及近红外光谱法分别进行烟叶化学成分检测,检测结果表明2种醇化方式的效果存在显著差异,真空包装方式对于烟碱的分解转化具有更强的阻隔作用,并且对醇化过程中还原糖及氯的降低有加速作用。     

连续流动分析法在烟草分析中的应用

连续流动分析法(CFA)作为一种快速、准确的批量检测方法已经广泛应用于国内外烟草行业。根据连续流动分析法的原理,详细介绍了该分析法在烟草中对烟碱、总糖及还原性糖、水溶性糖、总氮、蛋白质、氯以及钾等含量的测定方法及应用,并对连续流动分析法在烟草科学技术中的应用做了展望。     

Surface characteristics of antimony-doped tin oxide films

The surface composition of antimony-doped tin oxide films and its variation after electron bombardment and resistive heating were investigated. It was found that impurities such as carbon, potassium, sodium, calcium and chlorine segregate at the surface and that the carbon and chlorine are removed easily by electron bombardment. It is proposed that resistive heating caused the antimony to migrate rapidly to the surface at about 780°C, resulting in an antimony-rich phase which quickly reached a temperature high enough for sublimation. All substances except tin and oxygen and a trace amount of chlorine were removed. This process was accompanied by a resistivity change of the sample from 50 to 15000ω. Both electron bombardment and resistive heating above 780°C caused an increase in the tin-to- oxygen ratio as determined by Auger spectroscopy and electron spectroscopy for chemical analysis.     

Analysis of a potassium lidar system for upper-atmospheric wind-temperature measurements

We report a detailed analysis of wind-temperature (W/T) lidar systems based on mesospheric potassium as the tracer. Currently, most narrow-band (W/T) systems use sodium (Na) as the tracer because of its relatively large natural abundance, large cross section, and the ability to use Doppler-free Na spectroscopy to generate accurate absolute frequency markers. We show that a potassium-based system with existing near-infrared solid-state laser technology operating at the potassium D lines has the potential to make W/T measurements that are more accurate than current Na narrow-band systems and can be far simpler technically.     

Synthesis and optical properties of silver nanobars and nanorice

Silver nanobars with rectangular side facets and an average aspect ratio of 2.7 have been synthesized by modifying the concentration of bromide added to a polyol synthesis. Subsequent rounding of nanobars transformed them into nanorice. Due to their anisotropy, nanobars and nanorice exhibit two plasmon resonance peaks, scattering light both in the visible and in the near-infrared regions. With a combination of discrete-dipole approximation calculations and single-nanoparticle spectroscopy, we explored the effect of nanostructure aspect ratio and corner sharpness on the frequency of plasmon resonance. Near-field calculations and surface-enhanced Raman scattering measurements on single particles were performed to show how local field enhancement changes with both the wavelength and polarization of incident light.     

一种面向卷烟混合生产线的优化调度模型研究

目的 基于多品种、小批量的生产制造模式,在规定的8 h工作之内,快速有效地生产出多品种、多批量的卷烟,并对所需物料进行准确调度,达到经济效益最大化的目的。方法 针对某烟草生产企业订单需要生产6种型号的品牌香烟,通过分析卷烟生产线的工艺流程,提出一种解决卷烟厂车间资源优化调度的CSS模型,该模型可以根据产品之间的销售需求情况来匹配生产线资源配置,按需求比采用粒子群优化算法计算出单次投料后混合生产香烟所需的最小时间与最大收益。结果 将所得方案进行综合对比后,计算得出在规定工期内,生产香烟获取利益的最优分配方案,在迭代在10次以内时已完成了优化过程,最大获利为3.65万元。结论 该优化模型通过改变相关工艺参数能够实现对不同混合生产线的资源调度优化,并对其他制造行业提供借鉴价值。     

在线监测总糖、尼古丁的可行性研究

针对烟草加工中传统测试主要化学成分的方法，提出了“在线动态监测”这样一个全新的检测概念。并以总糖和尼古丁作为主要研究对象，就新方法应用的可行性进行了论证。实验数据表明，利用近红外仪器在线监测烟丝的总糖、尼古丁含量，具有较高的应用价值，为卷烟生产的动态监测提供了一种新的技术手段。     

Deep purification of detonation nanodiamond material

The paper presents new alternative procedures of chemical purification of detonation nanodiamonds and diamond-bearing detonation soot to remove water-insoluble metal-containing impurities through a high-temperature treatment using solutions of complexons of concentration 0.5 to 20 wt %, where the ratio between the detonation nanodiamond material and the complexon is above 0.2. The following substances can be used as complexons: sodium 2,3-dimercaptopropanesulfonate, disodium dihydrogen ethylenediaminetetraacetate (Trilon), thiocarbamide, potassium rhodanate, dicyandiamide, hexamethylenetetramine. Purification of detonation nanodiamonds can be also performed by exposing them to an ultrasonic action. A combination of the ultrasonic treatment and treatment with complexon solutions has proved most efficient, significantly reducing the amount of metal-containing impurities.     

Frequency-modulation spectroscopy with blue diode lasers

Frequency-modulation spectroscopy provides ultrasensitive absorption measurements. The technique is especially adaptable to diode lasers, which can be modulated easily, and has been used extensively in the near-infrared and infrared spectral regions. The availability of blue diode lasers now means that the accessible wavelength region can be increased. We successfully demonstrate wavelength-modulation spectroscopy and two-tone frequency-modulation spectroscopy for the weak second resonance line of potassium at 404.8 nm and for the transition at 405.8 nm in lead, starting from the thermally populated 6(p)(2 3)P(2) metastable level. Information on the modulation parameters is obtained with a fitting procedure. Experimental signal-to-noise ratios at different absorption levels are compared with theoretical signal-to-noise ratios and show good agreement. Detection sensitivities of 2 x 10(-6) and 5 x 10(-6) for wavelength and two-tone frequency-modulation spectroscopy, respectively, for a 120-Hz bandwidth are demonstrated.     

One-pot synthesis of near-infrared fluorescent gold clusters for cellular fluorescence lifetime imaging

A facile strategy to synthesize water-soluble fluorescent gold nanoclusters (Au NCs) stabilized with the bidentate ligand dihydrolipoic acid (DHLA) is reported. The DHLA-capped Au NCs are characterized by UV-vis absorption spectroscopy, fluorescence spectroscopy, transmission electron microscopy, and X-ray photoelectron spectroscopy. The Au NCs possess many attractive features including ultrasmall size, bright near-infrared luminescence, high colloidal stability, and good biocompatibility, making them promising imaging agents for biomedical and cellular imaging applications. Moreover, their long fluorescence lifetime (>100 ns) makes them attractive as labels in fluorescence lifetime imaging (FLIM) applications. As an example, the internalization of Au NCs by live HeLa cells is visualized using the FLIM technique.