基于神经网络的近红外光谱辛烷值模型的研究及应用

汽油管道调合的在线优化过程中,调合优化与控制系统对近红外光谱模型的依赖很大。光谱模型的精度及适应性直接影响整个在线调合系统。本文就如何建立适用于在线汽油调合的汽油辛烷值近红外光谱模型展开研究,提出一种采用主元分析(PCA)结合人工神经网络(ANN)的方法建立汽油近红外光谱辛烷值模型的方法;并与多元线性回归及偏最小二乘法建立的线性模型做比较。结果表明主元分析结合人工神经网络所建立的模型适应性较高、抗干扰能力强,适合汽油在线调合的现场应用。     

在线近红外光谱分析技术及应用

过程分析技术在优化生产控制、稳定产品质量、降低劳动成本、提高经济效益等方面起到越来越重要的作用,其中在线近红外光谱分析技术是目前发展最快和最具有前景的过程分析技术之一。介绍了在线近红外光谱分析系统的组成和特点,以及在石化、化工和制药领域的实际应用情况。应用实例表明,将在线近红外光谱分析技术用于生产过程可为厂家带来可观的经济效益。     

OPC技术在中药生产过程近红外智能监控系统中的应用

为解决近红外分析仪与自动控制系统通信的难题,提出一种运用OPC技术实现近红外在线分析系统与现场自动化设备进行数据双向读写的通信方法。该方法参考OPC技术规范,采用面向对象的设计思想,结合适配器模式与桥接模式封装OPC自动化接口,采用异步通信方式实现OPC客户端与OPC服务器的无缝通信,并成功应用于中药生产过程,实现提取等关键生产过程的近红外在线分析与智能监控。     

WinCC应用组态王XY曲线实现近红外在线分析

近红外光谱在线分析是一种最近广泛应用的过程分析技术。中药多成分近红外光谱在线分析需要作图控件在曲线显示方面具备多条光谱及其预测性质值的实时叠加显示、缩放、谱峰定位等功能,而广泛应用于在线监控的WinCC软件提供的作图控件不能满足需要。本文通过在WinCCV6.0软件中注册组态王超级XY曲线控件并应用VBSript语言编程实现,很好地解决了这一问题。将该技术应用于中药提取过程近红外光谱在线分析系统,使得操作员能够实时、直观地观察到近红外光谱及中药成分预测值的变化,有效保证生产工艺稳定。     

新型国产近红外分析仪的菜籽菜粕快速检测技术研究

近红外光谱分析作为1种快速无损的绿色分析技术,可应用于检测油料原料和油脂产品品质,为粮油企业带来巨大经济效益.本文采用一款新型国产化的近红外漫反射光谱分析仪对菜籽和菜粕进行了检测分析,分别对菜籽的水分、蛋白、含油量、脂肪酸、硫苷含量和菜粕水分、蛋白和含油量指标建立模型并验证.结果显示近红外检测结果与国标参考方法检测结果相关性好、准确度高.表明新型国产化的近红外分析仪可满足油脂企业对菜籽和菜粕产品的质量控制要求.     

近红外分析技术在汽油调合中的应用

近红外光谱分析是一种用于各种碳氢有机物组成和性质指标分析的重要检测技术.从近红外分析基本原理出发,分别介绍了近红外光谱的获取、传输、预处理、检测和分析技术的各种应用技术,探讨了定性和定量分析所涉及的各种分析、建模的步骤和方法,主要针对近红外光谱分析技术在汽油调合中的应用方法作了较为具体的说明.该技术已在炼化企业得到成功应用,并获得了明显的经济效益和社会效益.     

饲料蛋白固态发酵过程参数的FT-NIR快速检测

为实现对整个固态发酵过程状态信息变量的在线监测,研究提出基于近红外光谱分析技术的饲料蛋白固态发酵过程参数快速检测新方法。利用近红外光谱采集装置获取固态发酵物样本的近红外光谱,引入不同的光谱预处理方法校正原始近红外光谱数据,采用PLS建立固态发酵过程参数pH和湿度的定量分析模型。在模型校正过程中,通过交互验证法优选光谱预处理方法和建模所需的主因子数。试验结果显示,pH和湿度的PLS模型在预测集中的相关系数R_p分别为0.9838和0.8707,RMSEP分别为0.0696和0.0152。研究结果表明,利用近红外光谱分析技术快速监测固态发酵过程参数是可行的。     

亚麻籽油成分的近红外分析研究

亚麻油是重要的灰保健油。亚麻油中各种组分的定量分析，以往多采用气相色谱法，存在着方法繁琐、耗时较长的缺点，不适合生产时产品在线分析的需要。本文采用近红外光谱法（NIR）对亚麻油各种组分模型化后再进行定量分析，取得了满意的结果，基本达到了快速、无损、在线分析亚麻油的目的。     

基于PCA-SVM的柴油凝点近红外光谱软测量法

提出一种基于PCA-SVM及近红外光谱(NIRS)分析技术的柴油凝点软测量方法。首先,采用多项式卷积对原始的柴油NIRS数据进行光谱平滑、基线校正和标准归一化;然后,利用主元分析(PCA)对近红外光谱数据集的高维特征进行组合并向低维空间投影;最后,利用SVM回归算法建立凝点的软测量模型。与BP、SVM及PCA-BP方法相比,实验结果表明所提方法具有更高的测量精度,且与标准方法测量的结果更为接近,因此适合柴油凝点的在线测量。     

近红外光谱干涉仪高精度运动控制器设计

介绍一种以DSP TM320VC5402为核心的近红外光谱干涉仪中的音圈电机数字闭环控制方法.采用增量式数字PID控制策略,实现音圈电机匀速运动控制.该系统通过实验验证,达到速度偏差为满量程小于0.5%的控制精度要求,能够满足近红外光谱干涉仪动镜的精确匀速运动控制要求.     

苹果近红外光谱的特征提取和分类研究

采用近红外光谱分析法对不同种类的苹果样品进行分类,提出一种基于非相关判别转换的苹果近红外光谱定性分析新方法。实验分别采用主成分分析、Fisher判别分析和非相关判别转换三种方法对苹果光谱数据进行特征提取,并使用K-近邻分类算法建立三种苹果分类识别模型,最后使用"留一"交叉验证法进行模型检验。结果表明,使用非相关判别转换方法建立的模型正确识别率优于使用主成分分析和Fisher判别分析建立的模型。     

Extrapolation of the aerosol reflectance from the near-infrared to the visible: the single-scattering epsilon vs multiple-scattering epsilon method

In the atmospheric correction algorithm for Sea-viewing Wide Field-of-view Sensor, the effects of the spectral variation of the aerosol radiance contributions at the two near-infrared (NIR) bands are estimated directly from sensor-measured radiance. The aerosol effects at the NIR are then extrapolated into the visible through a process of aerosol model selection from evaluation of the NIR single-scattering epsilon value, which is defined as the ratio of the single-scattering aerosol reflectance between two NIR bands. The aerosol radiance contribution at the visible wavelengths is then removed. In this paper, a slightly different approach in the aerosol model selection and extrapolation, i.e. using the NIR multiple-scattering epsilon instead of the single-scattering epsilon, is examined. The NIR multiple-scattering epsilon is the ratio of the aerosol multiple-scattering reflectance between two NIR wavelengths. Simulations show that, in general, both methods give comparable results. Statistically, more than 95% of cases in the retrieved ocean colour spectrum are within required accuracy for both methods. For clear atmosphere, however, the results of the atmospheric correction using the single-scattering epsilon method usually performed slightly better than the multiple-scattering epsilon method. On the other hand, for the large aerosol optical thickness the multiple-scattering epsilon method has slightly better retrievals for the Tropospheric aerosols. Some detailed analyses and discussions are provided to explain differences in these two approaches for extrapolating and retrieving the aerosol effects in the visible.     

间歇硫酸盐法蒸煮过程纸浆卡伯值的近红外光谱法在线测量的研究

研究发现，硫酸盐法蒸煮液在800-900nm近红光谱波段间的吸收主要来自其中的碱木素，而其中的D-葡萄糖、D-甘露糖、D-木糖、D-半乳糖和糖醛酸等碳水化合物的降解产物及二氯甲烷和苯醇抽出物在此波段基本上没有吸收。并在这段波段建立起的纸浆卡伯值近红外光谱测量模型。该模型显示蒸煮液的吸光度与纸浆卡伯值具有较好的线性关系（相关系数R^2为0．996），并有较高的预测精度。因此，建议采用近红外光谱法实现硫酸盐法蒸煮过程纸浆卡伯值的在线测量，并有望开发出蒸煮过程纸浆卡伯值近红外光谱法在线测量仪，真正实现蒸煮过程的纸浆卡伯值的闭环反馈控制。     

近红外和高光谱技术用于小麦籽粒蛋白含量预测探索

以57份小麦籽粒为样本,分别采用Vertex 70傅立叶近红外光谱仪(单籽粒测样附件)和近红外增强型高光谱成像系统采集光谱,建立预测小麦籽粒蛋白含量的模型。基于近红外的小麦单籽粒模型相关系数为0.52,交叉校验均方根误差为0.807;而基于高光谱建立的模型相关系数为0.81,交叉校验均方根误差为0.7035。结果表明:在样本量少且为籽粒形态时,可优先考虑高光谱技术替代传统的近红外单籽粒采样模式来检测样本内部品质,但其实用性还有待进一步验证。若深入结合图像信息,高光谱技术在农产品内外品质检测方面有更广阔的应用。     

Improvement of the visibility of concealed features in artwork NIR reflectograms by information separation

Near Infrared (NIR) reflectography, coupled to visible (VIS) one, is a spectrophotometric imaging technique employed to probe both the inner and the outer layers of artworks. NIR reflectograms may partially contain information pertinent to the visible spectrum (due to the poor pigment transparency in NIR) and this decreases their comprehensibility. This work presents an innovative digital processing methodology for accentuating information contained in the infrared reflectograms. The proposed method consists of inducing minor changes in pixel intensity by suppressing VIS information content from NIR information content. The method creates such enhanced NIR reflectogram by extrapolating VIS reflectogram to a reflectogram recorded in NIR range and by subtracting it from the measured values in the near infrared spectral sub-band. As an extrapolator we suggest a feed forward artificial neural network (ANN). Significant results of improved visualization are exemplified on reflectograms acquired with a VIS-NIR $〈400,2250〉\text{nm}$ scanning device on real paintings such as Madonna dei Fusi attributed to Leonardo da Vinci. Parameters of the method, artificial neural network and separability of used pigments are discussed.     

连续小波变换-支持向量回归用于植物样品多组分分析

采用连续小波变换（CWT）技术对近红外光谱（NIR）数据进行预处理,扣除光谱中的背景与噪音成分,再用支持向量回归（SVR）进行建模,建立了用于复杂植物样品多组分分析的建模方法（CWT-SVR）,并应用于烟草样品中常规成分（总糖、总植物碱和总氮）含量的测定。结果表明,CWT—SVR方法优于基于全谱数据的SVR和偏最小二乘（PLS）法,为近红外光谱定量分析提供了一种新的建模方法。     

Quantitative and qualitative analyses in near infrared analysis of basic compounds in sugar beet leaf

The application of the near infrared technique (NIR) on a large scale is possible thanks to the use of rapid and inexpensive computers, and facilitates both quantitative and qualitative analyses. The great advantage of NIR is rapid analyses and its non-destructive character, and possibility of measuring many different components at the same time. The NIR method was used to study changes in chemical composition in leaves of sugar-beet under the influence of various fertilizers. The results obtained show that this method is an effective tool to predict protein, nitrogen and saccharides.     

Evaluation of MODIS SWIR and NIR-SWIR atmospheric correction algorithms using SeaBASS data

Using the NASA maintained ocean optical and biological in situ data that were collected during 2002-2005, we have evaluated the performance of atmospheric correction algorithms for the ocean color products from the Moderate Resolution Imaging Spectroradiometer (MODIS) on Aqua. Specifically, algorithms using the MODIS shortwave infrared (SWIR) bands and an approach using the near-infrared (NIR) and SWIR combined method are evaluated, compared to the match-up results from the NASA standard algorithm (using the NIR bands). The in situ data for the match-up analyses were collected mostly from non-turbid ocean waters. It is critical to assess and understand the algorithm performance for deriving MODIS ocean color products, providing science and user communities with the important data quality information. Results show that, although the SWIR method for data processing has generally reduced the bias errors, the noise errors are increased due mainly to significantly lower sensor signal-noise ratio (SNR) values for the MODIS SWIR bands, as well as the increased uncertainties using the SWIR method for the atmospheric correction. This has further demonstrated that future ocean color satellite sensors will require significantly improved sensor SNR performance for the SWIR bands. The NIR-SWIR combined method, for which the non-turbid and turbid ocean waters are processed using the NIR and SWIR method, respectively, has been shown to produce improved ocean color products.     

A simple ‘clear water’ atmospheric correction algorithm for Landsat-5 sensors. I: A spectral slope-based method

An atmospheric correction algorithm (the improved ‘clear water’ atmospheric correction algorithm for broadband (ICAB) sensors, or ‘ICAB’ algorithm) was constructed and tested, and the results are presented in this article. The Rayleigh scattering contributions at three visible bands and one near-infrared (NIR) band were estimated using the look-up table method. The remote-sensing reflectance associated with aerosol scattering at the NIR band was estimated using the ‘clear water’ method. The Ångström exponential model was used to extrapolate the aerosol scattering contribution from the NIR bands into visible bands. The spectral slopes of remote-sensing reflectance between 485 and 560 nm were developed for estimating the exponential coefficient of the Ångstrom exponential model. According to the case study performed in Taihu Lake, China, the results of which are presented in this article, it has been found that the spectral characteristics in shape and magnitude estimated by the ICAB algorithm are quite similar to the typical spectral characteristics of water bodies. The uncertainties of the ICAB algorithm were 7.45%, 10.57%, 13.51%, and 17.14%, respectively, at 485, 560, 660, and 830 nm.     

近红外光谱数据分析方法研究

文中将主成分分析和BP神经网络方法相结合,用于对近红外光谱数据进行预处理和回归分析,较好地解决了近红外分析中的非线性关联问题。实验结果表明,该方法在近红外光谱数据的分析中与传统的化学计量学方法相比有较好的应用效果。