小波神经网络在紫外光谱识别中的应用研究

介绍用于信号识别的小波神经网络的结构和算法，并将其用于酪氨酸，二羟基苯丙氨酸和色氨酸的紫外光谱识别。在波小神经网络中，采用Ｍｏｒｌｅｔ母波波和一维搜索变步长共轭梯度优化方法。结果表明，小波神经网络对于光谱间的细微结构差别具有很好的识别能力。     

维生素B的同时测定——神经网络与紫外光谱法

研究了反向传播神经网络（ＢＰＮＮ）及改进算法，经比较发现ＢＰＮＮ在某些方面优于滤波分辨及多元校正，将其应用于多元分析及光谱定量，可不经分离同时测定Ｂ族维生素，操作简便，结果良好，从而开发一种多组分同时光谱测定新方法。     

紫外光谱多元线性回归法同时测定硝基苯与亚硝基苯和苯胺

利用硝基苯，亚硝基苯和苯胺紫外吸收光谱的差异，用多元线性回归法处理多波长下的测量数据，实现了三发的同时准确测定。     

维生素B的同时测定─—神经网络与紫外光谱法

研究了反向传播神经网络（ＢＰＮＮ）及改进算法，经比较发现ＢＰＮＮ在某些方面优于滤波分辨及多元校正。将其应用于多元分析及光谱定量，可不经分离同时测定Ｂ族维生素，操作简便，结果良好。从而开发了一种多组分同时光谱测定新方法。     

人工神经网络—紫外光谱定量多组分体系的研究

本文系统地考察了人工神经网络（ＡＮＮ）－紫外光谱（ＵＶＳ）同时定量多组分混合溶液时参数选择时网络训练和预报性能的影响。合理选取诸参数，可提高训练效率改善预报性能，而且所优化的参数集可移植到其它相似体系。     

不定物种的多组分体系同时测定：逐步回归法解析紫外光谱矢量数据的研究

本文采用逐步回归法对不定物种的多组分体系的紫外光谱矢量数据进行解析，可对物种及物种含量同时进行定性和定量分析，通过对两个复方药物体系（复方扑热息痛与复方维生素Ｂ）的紫外光谱数据的实际解析，表明本法能有效地完成混合物中物种及物种含量的同时定性和定量测定。     

人工神经网络用于紫外光谱同时测定苯和甲苯及二甲苯的含量

根据反向传输（Backpropagation,BP）算法,应用三层ANN（人工神经网络,Artificial neural network）网络原理,对紫外光谱严重重叠的苯、甲苯和二甲苯的混合体系进行同时测定。在230～280nm范围内,以16个特征波长处的紫外吸光度作为网络特征参数,并通过均匀设计安排样本进行网络训练和计算。苯、甲苯和二甲苯的回收率依次为98．7％,99．4％和97．4％,测定结果的相对标准偏差分别为2．0％,2．8％和2．7％。     

用紫外光谱法验证磁化杯对水溶液的作用

水接受磁场能后,其物理化学性能会发生某些变化,提高了水的“生物活性”。此时的水被称为磁处理水。磁处理水具有某些保健功效已有报道。时下,磁化杯正畅销市场。国内外学者采用多种物理化学测试手段研究磁处理水,紫外吸收光谱法是其中之一。可是,磁化杯对水作用后所引起水的性能变化,往往不易被探测,测得的结果常常难于重复。作者利用葡萄糖水溶液具有的紫外吸收光谱特征峰(在280nm处),验证磁化杯对水溶液的作用。     

反相高效液相色谱与紫外光谱对硝基苯类化合物的联合定性

结合高效液相色谱和紫外光谱各自的优点和信息，建立了有机化合物的高效液相色谱和紫外光谱联合定性方法。在高效液相色谱中，二极管阵列检测器实时在线地提供样品组分的紫外光谱，由此建立紫外光谱谱图库，并发展了谱图库检索算法，同时测定了硝基苯类化合物的反相色谱用热力学参数ａ、ｃ值。结果表明，该法是可行的，满足定性所需的精度。     

Simultaneous Quantitation of Three Aromatic Compounds by Fluorescence Spectroscopy and Elastic Component Regression

The determination of aromatic compounds is of importance in many fields. Even if these compounds may be determined by molecular fluorescence and second-order calibration, the methods are relatively complex and require higher mathematics. Here is explored whether the reduction of the dimensions, i.e., transforming two-dimensional data to one-dimensional data followed by a special variable selection and calibration process, is feasible for this purpose. The present work investigated the feasibility of combining competitive adaptive reweighted sampling and elastic component regression for the direct evaluation of the spectra to quantify three components simultaneously. Model population analysis has also been performed. The results confirmed the feasibility of the procedure, which may serve as an alternative method to simultaneously determine aromatics.     

人工神经网络紫外分光光度法同时测定去痛片组分含量

用BP人神经网络处理复方制剂去痛片的紫外吸收光谱数据,达到了对其各组分含量进行同时测定的目的。通过对网络结构和参数的优化,提高了预报的准确度。     

用人工神经网络-近红外光谱法测定冬虫夏草中的甘露醇

提出了用近红外漫反射光谱技术快速分析发酵冬虫夏草菌粉中甘露醇含量的新方法。采用比色法测定样品中的甘露醇，其含量范围为8．082％-14．548％。在7501．7-6097．8cm^-1与5453．7-4246．5cm^-1波段，分别采用PCR、PLSR和BP神经网络方法建立了样品近红外光谱的一阶微分光谱与其甘露醇含量之间的相关模型。BP神经网络模型的内部交叉验证误差均方根为0．475，预测误差均方根为0．608，均优于PCR和PLSR的处理结果。这表明，BP神经网络法对非线性检测对象具有较好的建模效果，可用于中药近红外光谱分析的非线性校正。     

紫外分光光度法同时测定硝酸盐氮和亚硝酸盐氮

提出了用均匀设计安排ＢＰ网络运行参数以进行优化的方法。研究了硝酸盐和亚硝酸盐在紫外区的吸收光谱。利用ＢＰ算法的人工神经网络处理光谱数据,提出了关于同时测定水样中的硝酸盐氮和亚硝酸盐氮的新方法。对实际水样进行了测定,结果满意。     

人工神经网络分光光度法同时测定钼和铬

在钼，铬－二溴邻硝基苯基荧光酮－ＣＴＭＡＢ显色体系中，应用三层ＡＮＮ－ＢＰ网络解析钼和铬的吸收光谱，不经分离分光光度法同时测定钼和铬。钼和铬的表观摩尔吸光系数分别为ε５４５＝ ６．４８×１０４Ｌ·ｍ ｏｌ－ １·ｃｍ － １，ε５４７＝ １．５４×１０４Ｌ·ｍ ｏｌ－ １·ｃｍ － １。对合成样品和合金钢中钼和铬进行了同时测定，结果满意。使用改进的ＢＰ算法，避免了神经网络学习过程中可能产生的麻痹现象。提出了目标向量的简单变换方法及便于网络参数选择的收敛评价函数。     

神经网络反传改进算法及其用于分析复方药物

提出了一种反向传播改进算法ＭＢＰ，能加快神经网络ＮＮ的收敛速度，增强反传算法的实用性。用于分析多元复方药物，结果良好，操作简便。     

神经网络Kalman滤波算法及多组分光度分析应用

前馈神经网络ＮＮ误差反向传播算法（ＢＰ）收敛速度较慢且常陷入局部极优值等，针对此种缺陷提出了一种基于扩展Ｋａｌｍａｎ滤波的快速学习新算法（ＥＦ）。与ＢＰ相比，ＥＦ法不仅学习效率高收敛速度快，数值稳定性好，而且所需学习次数少，调节参数少，由非线性系统建模与辨识的模拟结果表明，ＥＦ是提高网络收敛速度改善神经学习性能的一种有效方法，谈谈用于多组分光谱分析，结果良好。     

The use of Artificial Neural Networks for the selective detection of two organophosphate insecticides: Chlorpyrifos and chlorfenvinfos

Amperometric acetylcholinesterase (AChE) biosensors have been developed to resolve mixtures of chlorpyrifos oxon (CPO) and chlorfenvinfos (CFV) pesticides. Three different biosensors were built using the wild type from electric eel (EE), the genetically modified Drosophila melanogaster AChE B394 and B394 co-immobilized with a phosphotriesterase (PTE). Artificial Neural Networks (ANNs) were used to model the combined response of the two pesticides. Specifically two different ANNs were constructed. The first one was used to model the combined response of B394 + PTE and EE biosensors and was applied when the concentration of CPO was high and the other, modelling the combined response of B394 + PTE and B394 biosensors, was applied with low concentrations of CPO. In both cases, good prediction ability was obtained with correlation coefficients better than 0.986 when the obtained values were compared with those expected for a set of six external test samples not used for training.     

Determination of Active Components in a Natural Herb with Near Infrared Spectroscopy Based on Artificial Neural Networks

The non-linear relationships between the contents of ginsenoside Rg1, Rb2, Rd and Panax notoginseng saponins(PNS) in Panax notoginseng root herb and the near infrared(NIR) diffuse reflectance spectra of the herb were established by means of artificial neural networks(ANNs). Four three-layered perception feed-for-ward networks were trained with an error back-propagation algorithm. The significant principal components of the NIR spectral data matrix were utilized as the input of the networks. The networks architecture and parameters were selected so as to offer less prediction errors. Relative prediction errors for Rg1, Rb1, Rd and PNS obtained with the optimum ANN models were 8.99%, 6.54%, 8.29%, and 5.17%, respectively, which were superior to those obtained with PLSR methods. It is verified that ANN is a suitable approach to model this complex non-linearity. The developed method is fast, non-destructive and accurate and it provides a new efficient approach for determining the active components in the complex system of natural herbs.