基于高光谱技术的玉石鉴定及模型研究

为了给玉石鉴定提供依据以及得到优化预测模型,分别对天然玉石和假玉石的可见光高光谱图像进行分析。针对高光谱图像数据的非线性、小样本以及空间光谱维数大等问题,本研究首先对原始光谱数据进行主成分分析(PCA),使高维光谱数据降维,通过对比分析其平均光谱图和方差贡献率图,发现天然玉石与假玉石的谱线之间存在很大的差距,证明了高光谱成像技术在玉石鉴定领域的可行性。然后分别采用费希尔(Fisher)判别法、反向传输(BP)神经网络以及支持向量机(SVM)判别法建立的三种数学模型对玉石进行分类模式判别,结果显示,用Fisher判别法能直接得到预测的类别归属,用BP神经网络以及SVM判别法得到的类别鉴定准确率分别为96.37%,82.5%。研究结果表明,高光谱技术结合BP人工神经网络预测建模方法可以作为快速和非破坏性预测玉石真假的有效手段。     

大鼠胰腺及癌组织红外光谱连续小波特征提取及径向基人工神经网络识别

采用水平衰减全反射(HATR)傅里叶变换红外光谱法(FTIR)测定了SD大鼠胰腺正常组织与非正常组织的谱图,提出了一种新的基于FTIR的连续小波特征提取与径向基人工神经网络分类方法以提高FTIR对早期SD大鼠胰腺癌的诊断准确率。利用连续小波多分辨率分析法提取FTIR特征量,对于提取的特征量采用径向基函数神经网络进行模式分类。对SD大鼠的胰腺正常组织、早期癌组织及进展期癌组织的FTIR,利用连续小波多分辨率分析法提取9个特征量,进行RBF神经网络分类判断。当目标误差为0.01,径向基函数的分布常数为5时,网络达到最优化,总的正确识别率为96.67%。并对影响分类结果的网络参数、目标误差和分布常数对分类样品的影响做了讨论。实验结果表明:此方法对早期胰腺癌具有较高的诊断率。     

基于料酒的三维同步荧光光谱模式识别研究

提出了一种应用同步荧光光谱技术无损快速鉴别料酒品牌的新方法.利用主成分分解法和小波变换法对料酒样品的同步荧光光谱信号进行了压缩处理,分别采用同步荧光光谱数据的第一主成分和小波细节系数为特征变量进行主成分分析和聚类分析,分类结果表明小波系数作为料酒的特征变量对料酒品牌分类正确率更高.利用偏最小二乘和径向基人工神经网络方法...     

人工神经网络方法用于脉冲极谱重叠峰解析

将人工神经网络用于脉冲极谱法中Ｐｂ和Ｔｌ重叠信号峰的解析，对神经网络参数的影响及优化作了研究。结果表明，网络的增益，学习速率和动量是影响网络收敛和稳定性的关键参数。本文还将神经网络与偏最小二乘法的计算结果作了比较。     

稀土—氮复合孕育剂对灰铸铁石墨组织的影响

采用主碳当量灰铸铁作为试验材料，以孕育处理的方式加入稀土和氮，研究了稀土，氮及二者复合能对石墨数量和石墨表面形貌的影响。试验结果表明，稀土和氮，特别是它们复合加入，使石墨百分含量减少；氮和氮与稀土复合孕育使灰铸铁的石墨表面变粗糙，且端部有钝化现象。     

基于自组织映射神经网络的中药注射剂质量快速鉴别方法

将近红外光谱分析技术与人工神经网络相结合,研究提出一种基于自组织映射神经网络的近红外光谱神经元分类模型,用于对中药注射剂产品的近红外光谱进行计算分析,可实现对注射剂质量的快速鉴别。以3个不同厂家生产的参麦注射剂为研究对象,考察本方法的分类能力,其分类正确率达到96.4%,优于参与比较的判别式偏最小二乘法(90.5%)、反向传播神经网络(88.1%)和支持向量机(90.5%)。     

人工神经网络-近红外光谱法非破坏监测芦丁药品的质量

用近红外漫反射光谱法非破坏监测芦丁药品的质量。利用人工神经网络化学计量学分类技术，建立三层神经元的神经网络，对网络参数进行优化选择以建立最佳网络模型。根据芦丁药品的近红外漫反射光谱，成功地分辨出合格药品和不合格药品。     

稀土在铸铁中的变质行为

本文简述了铸铁的发展史,指出孕育是铸铁发展的里程碑。作者认为,孕育的实质是增加铸铁的共晶团数,而共晶团数则与铸铁的一定石墨形态相对应。为此,作者提出了共晶团数与铸铁的石墨形态以及抗拉强度对应关系的示意图。 作者把稀土在铸铁中的主要作用归结为变质作用。根据国内外的大量研究,作者着重指出,稀士在球墨铸铁、蠕墨铸铁和灰铸铁中具有突出的生核作用。只要加入量适当,稀土在铸铁中的变质行为是富有成效的,是独特的,是任何其它元素所不能取代的。因此,作者认为,稀土在生产球墨铸铁、蠕墨铸铁和高强度灰铸铁中具有广阔的应用前景。     

化学中的人工神经网络法

反向传输人工神经网络是应用最为广泛的一种方法, 本文较详细地介绍了该种方法及其相关的问题, 同时给出了Kohonen 模型和Hopfield 网络的简单算法。关于神经网络在化学中的应用, 该文介绍了6 个方面: 定量结构2活性性质相关性(QSAR/QSPR )研究, 有机化合物结构解析, 光谱的数据处理, 化学反应性, 流程优化, 故障诊断及控制, 蛋白质结构。     

稀土在灰口铸铁各相中的含量及其显微分布的研究

应用有机电解液低温电解和放射性同位素及等离子光谱分析方法,测定了经稀土处理的三种灰口铸铁(暖气片、铸铁管和钢锭模)的合金化稀土量及石墨中稀土量,结果表明,合金化稀土量随稀土总量增加而增加,随硫含量增加而降低;石墨中稀土量远高于合金化稀土量,一般也高于稀土总量。涂敷法自射线照像结果表明,灰铸铁及蠕墨铸铁的石墨中富集有稀土,初生石墨比共晶石墨富集有更多的稀土;在初生石墨中,稀土的分布是心部高于边缘,在星形石墨内及其周围都有大量稀土富集。电子探针分析与化学分析表明,石墨中同时存在有多种元素的吸附;此外,发现稀土对灰口铸铁有细化晶粒团及脱氧作用。     

Automatic classification of graphite in cast iron.

A method for automatic classification of the shape of graphite particles in cast iron is proposed. In a typical supervised classification procedure, the standard charts from the ISO-945 standard are used as a training and validation population. Several shape and size parameters are described and used as discriminants. A new parameter, the average internal angle, is proposed and is shown to be relevant for accurate classification. The ideal parameter sets are determined, leading to validation success rates above 90%. The classifier is then applied to real cast iron samples and provides results that are consistent with visual examination. The method provides classification results per particle, different from the traditional per field chart comparison methods. The full procedure can run automatically without user interference.     

Application of Back-propagation Artificial Neural Network in Speciation of Cadmium

A method for predicting the five species contents of cadmium was developed by combining the back-propagation artificial neural network with graphite furnace atomic absorption spectrometry(BP-ANN-GF-AAS).Based on the strong learning function and the features of the information distributed storage of artificial neural network(ANN),a single ANN was constituted in which only one determination point of every sample was required.The exchangeable,carbonated,Fe-Mn oxidable,organic and residual species of cadmium for 20 kinds of soil samples from the two sections of Changchun(China) were determined by BP-ANN-GF-AAS.The detection limit of the method is 0.024 μg/L and the limit of quantification is 0.080 μg/L.t-Test indicates that there is not any systemic error of the results obtained by the Tessier sequential extraction graphite furnace atomic absorption spectrometry method(Tessier-GF-AAS) and BP-ANN-GF-AAS.Compared with those of the Tessier-GF-AAS,the prediction errors of BP-ANN-GF-AAS are less than 10%.The proposed method is fast,convenient,sensitive,and can eliminate the interference among various species.     

稀土变质及热处理对耐磨铸铁冲击疲劳性能的影响

采用金相显微镜、扫描电镜观察了经冲击疲劳试验后耐磨铸铁中碳化物的形貌、疲劳裂纹的萌生与扩展,测定了稀土含量及加热温度与裂纹的长度和裂纹扩展之间的关系曲线,在此基础上探讨了稀土变质及热处理对耐磨铸铁冲击疲劳性能的影响.结果表明: 稀土能推迟裂纹萌生的时间,降低裂纹扩展速率,提高其冲击疲劳抗力.当稀土与热处理共同作用时,效果更显著.其原因主要归于网状共晶碳化物形态与分布的改变.     

Principal Component‐Wavelet Neural Network as a Multivariate Calibration Method for Simultaneous Determination of Iron,Nickel, and Cobalt

Abstract

Principal component‐artificial neural network (PC‐ANN) and principal component‐wavelet neural network (PC‐WNN) are applied for simultaneous determination of iron(II), nickel(II), and cobalt(II). A simple and selective spectrophotometric method for simultaneous determination of iron(II), nickel(II), and cobalt(II) based on formation of their complexes with 1‐(2‐pyridylazo)‐2‐naphtol (PAN) in micellar media is described. Although the complexes of Fe(II), Ni(II), and Co(II) with reagent show a spectral overlap, they have been simultaneously determined by PC‐ANN and PC‐WNN. The results obtained by the two methods were compared and it was shown that in PC‐WNN, the convergence speed was faster and the root mean square error of prediction set was also smaller than PC‐ANN. Interference effects of common anions and cations were studied and the proposed method was also applied satisfactorily to the determination of Fe(II), Ni(II) and Co(II) in synthetic samples.     

Classification of time-of-flight secondary ion mass spectrometry spectra from complex Cu–Fe sulphides by principal component analysis and artificial neural networks

Artificial neural network (ANN) and a hybrid principal component analysis-artificial neural network (PCA-ANN) classifiers have been successfully implemented for classification of static time-of-flight secondary ion mass spectrometry (ToF-SIMS) mass spectra collected from complex Cu–Fe sulphides (chalcopyrite, bornite, chalcocite and pyrite) at different flotation conditions. ANNs are very good pattern classifiers because of: their ability to learn and generalise patterns that are not linearly separable; their fault and noise tolerance capability; and high parallelism. In the first approach, fragments from the whole ToF-SIMS spectrum were used as input to the ANN, the model yielded high overall correct classification rates of 100% for feed samples, 88% for conditioned feed samples and 91% for Eh modified samples. In the second approach, the hybrid pattern classifier PCA-ANN was integrated. PCA is a very effective multivariate data analysis tool applied to enhance species features and reduce data dimensionality. Principal component (PC) scores which accounted for 95% of the raw spectral data variance, were used as input to the ANN, the model yielded high overall correct classification rates of 88% for conditioned feed samples and 95% for Eh modified samples.     

Elimination of iron interference in the molecular spectrophotometric determination of aluminum in soil extracts using artificial neural networks

An artificial neural network (ANN) calibration model was developed to determine aluminum in the presence of iron in soil extracts, using xylenol orange as chromogenic reagent. The spectral data of synthetic mixtures of Al(3+) and Fe(3+) as well as of the soil extracts, were recorded in the range between 410 and 580 nm. Method validation was carried out using 18 soil extracts. The results gave good linear correlations between the ANN model and the ICP OES measurements for both species.     

Detection of preclinical scrapie from serum by infrared spectroscopy and chemometrics

In this study we describe a methodology for diagnosing preclinical scrapie infection in hamsters from serum by a combination of Fourier-transform infrared (FT-IR) spectroscopy and chemometrics. Syrian hamsters (Mesocricetus auratus) were orally inoculated with the 263K scrapie agent, or mock-infected, and sera were obtained at 70, 100 and 130 days post infection (dpi) and at the terminal stage of scrapie (160 +/- 10 dpi). The analysis of hamster sera by FT-IR spectroscopy and artificial neural networks (ANN) confirmed results from earlier studies which had indicated the existence of disease-related structural and compositional alterations in the sera of infected donors in the terminal stage of scrapie [Schmitt et al. (2002) Anal Chem 74:3865-3868]. Experimental data from sera of animals in the preclinical stages of scrapie revealed subtle but reproducible spectral variations that permitted the identification of a preclinical scrapie infection at 100 dpi and later, but not at 70 dpi. The IR spectral features that were discriminatory for the preclinical stages differed from those of the terminally ill individuals. In order to reliably identify scrapie-negative as well as preclinical (100 and 130 dpi) and terminal scrapie-positive animals, a hierarchical classification system of independent artificial neural networks (ANN) was established. A "toplevel" ANN was designed which discriminates between animals in the terminal stage of scrapie and preclinical scrapie-positive or control animals. Spectra identified by the "toplevel" ANN as preclinical or controls were then further analyzed by a second classifier, the "sublevel" ANN. Using independent external validation procedures, the toplevel classifier produced an overall classification accuracy of 98%, while the sublevel classifier yielded an accuracy of 93%, indicating that scrapie-specific serum markers were also present for preclinical disease stages. Possible spectral markers responsible for the discrimination capacity of the two different ANNs are discussed.