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。     

三维荧光光谱总体积积分法同时测定多环芳烃

提出了三维荧光光谱总体积积分法并应用于多环芳烃萘、芘、Ｂｅｉ的同时测定，结果表明，三维荧光光谱总体积积分法不仅可以同时测定萘、芘、Ｂｅｉ而且灵敏度（信噪经）常规法分别提高２２、３３和２６倍同时并考察了积分区域对灵敏度的影响。     

导数—同步荧光光谱法同时测定三种B族维生素的研究

研究了维生素Ｂ１的氧化产物（硫色素）、维生素Ｂ２和Ｂ６混合物体系的导数－同步荧光光度谱，提出了混合物体系荧光光谱被分辨开的同时测定方法，该法可不经分离直接测定复合维生素片剂中维生素Ｂ１，Ｂ２及Ｂ６，其线性范围均为０－１．０μｇ／ｍＬ，检出限分别为０．５、１．５及４．０ｎｇ／ｍＬ。     

三维荧光光谱总体积积分法同时测定维生素B2和B6

运用三维荧光光谱总体积积分法建立了一种测定维生素Ｂ２和Ｂ６的荧光分析新方法，该方法的度较常规峰值法分别提高１３０和３１０倍，同时成功地应用于两者的同时测定。     

多元非线性荧光校正的人工神经网络方法

由实验测定得知罗丹明Ｂ，丁基罗丹明Ｂ，曙红Ｂ组成的三组分混和荧光分析体系存在严重的荧光熄灭现象，其混和物的荧光光谱呈非线性，ＰＬＳ难以校正。本文成功地将ＢＰ－ＡＮＮ应用于此多元非线性荧光校正问题，完成了三组分的同时测定，得到了满意的结果。     

卡尔曼滤波分光光度法同时测定水中微量砷和磷

研究了砷钼酸和磷钼酸与乙基罗丹明Ｂ的缔合物的吸收光谱，并以卡尔曼滤波算法对这两个严重重叠的光谱进行数据解析，达到了不需分离和掩蔽而同时测定砷和磷的目的，本法准确，精密，应用于水样分析，获得满意结果。     

比光谱导数分光光度法同时测定微量磷和砷

采用比光谱导数法对严重重叠的砷钼酸和磷钼酸与乙基罗丹明Ｂ的缔合物的吸收光谱进行数据解释。结果表明，此法不需分离和掩蔽而能实现对微量砷和磷的同时测定，探讨了显色条件，测定波长及求导计算时波长间距的选择等问题，应用于合成样品和水样的分析。     

偏最小二乘法用于同步荧光法同时测定维生素B1,维生素B2和维生素B6

本文将偏最小二乘法结合同步荧光扫描技术对含维生素Ｂ１，Ｂ２和Ｂ６的混合物进行了同时测定。对同步荧光法的测定条件及△λ的选择进行了试验和讨论，比较了△λ分别为６５ｎｍ和３０ｎｍ时的计算结果。所建立的方法用于复合维生素Ｂ片等药片叶Ｂ１，Ｂ２和Ｂ６孤同时测定，获较满意的结果。     

湖泊水体中土霉味化合物2-甲基异莰醇的合成与测定

利用甲基格氏试剂和天然樟脑反应，合成土霉味化合物２－甲基异莰醇（ＭＩＢ），并用色谱、质谱和红外光谱对其进行鉴定，然后以合成ＭＩＢ作标样，测定ＭＩＢ的气相色谱保留值及其在东湖水体中浓度随季节变化规律。研究结果表明，东湖水体产生土霉味的主要成份为ＭＩＢ，在１９９５～１９９６年期间，１月、４月和９月为武汉东湖水体中ＭＩＢ浓度的高峰期，经测定东湖水中ＭＩＢ的ＧＣ保留值与合成的ＭＩＢ的ＧＣ保留值完全一致。     

滇姜花新二萜成分的研究

从滇姜花中分离并鉴定了５个二萜成分，其中新化合物滇姜花素Ａ，Ｂ的结构经光谱及化学方法推定。     

神经网络扩展滤波算法及其多元光谱分辨应用

前馈神经网络(NN)误差反向传播算法(BP)应用较广, 但收敛较慢且易陷入局部极优, 针对这一不足, 本文提出了一种基于扩展滤波的快速学习新颖算法(EF)。与BP相比较, 该法不仅具有学习效率高, 收敛速度快, 所需学习次数少,数值稳定性好等优点, 而且所需调节参数少。由非线性系统建模与辨识的模拟结果表明, EF是一种有效的神经学习新算法。该法用于多元光谱校正与分辨,获得良好结果。     

神经网络法用于多元混酸同时测定

利用多层神经网络误差反向传播算法处理酸碱电位滴定数据，求得出多元混合酸各组分的浓度，优化了神经网络的结构和参数，测定了三组分有机酸混合样品，结果良好，平均相对偏差ＲＳＤ≤４％。     

Prediction of Rheological Properties of Multi-Component Dispersions by Using Artificial Neural Networks

This article shows the ability of artificial neural network (ANN) technology for predicting the correlation between rheological properties of multi-component food model systems and their chemical compositions. Multi-component food model systems were made of whey protein isolate (WPI) (2, 4 wt%), Iranian tragacanth gum (TG) (Astragalus gossypinus) (0.5, 1 wt%) and oleic acid (5, 10% v/v). The input parameters of the neural networks (NN) were these chemical compositions, namely WPI and TG concentrations, and oleic acid volume fractions. The output parameters of the NN models were rheological properties of multi-component food model systems (flow and consistency indices, viscosity, loss and storage moduli). Results showed that, ANN with training algorithm of back propagation (BP) was the best one for the creation of nonlinear mapping between input and output parameters. The best topology was 3-10-5. The ANN model predicted the rheological properties of multi-component food model systems with average RMSE 4.529 and average MAE 3.018. These results show that the ANN can potentially be used to estimate rheological parameters of multi-component food model systems from chemical composition. This development may have significant potential to improve product quality control and reduce time and costs by minimizing the rheological experiments.     

Accelerating the Construction of Neural Network Potential Energy Surfaces: A Fast Hybrid Training Algorithm

Machine learning approaches have been promising in constructing high-dimensional potential energy surfaces (PESs) for molecules and materials. Neural networks (NNs) are one of the most popular such tools because of its simplicity and efficiency. The training algorithm for NNs becomes essential to achieve a fast and accurate fit with numerous data. The Levenberg-Marquardt (LM) algorithm has been recognized as one of the fastest and robust algorithms to train medium sized NNs and widely applied in recent NN based high quality PESs. However, when the number of ab initio data becomes large, the efficiency of LM is limited, making the training time consuming. Extreme learning machine (ELM) is a recently proposed algorithm which determines the weights and biases of a single hidden layer NN by a linear solution and is thus extremely fast. It, however, does not produce sufficiently small fitting error because of its random nature. Taking advantages of both algorithms, we report a generalized hybrid algorithm in training multilayer NNs. Tests on H+H₂ and CH₄+Ni(111) systems demonstrate the much higher efficiency of this hybrid algorithm (ELM-LM) over the original LM. We expect that ELM-LM will find its widespread applications in building up high-dimensional NN based PESs.     

电化学衍生-高效液相色谱和荧光光度法检测花生酱中的黄曲霉毒素

建立了免疫亲和柱净化-柱后电化学衍生-高效液相色谱结合荧光光度法检测花生酱中4种黄曲霉毒素(B1、B2、G1和G2)的方法。样品经过体积分数为60%的甲醇提取,通过免疫亲和柱净化后,以KobraCell装置柱后衍生,高效液相色谱法分离定量。黄曲霉毒素B1、B2、G1和G2能达到完全的基线分离,检测限分别为0.5、0.15、0.5和0.15μg/kg,线性相关系数0.999,回收率可达74.2%～96.5%,相对标准偏差低于11%。该方法能够满足花生酱中黄曲霉毒素检测的需要。     

掺硼纳米TiO_2对农药毒死蜱的光催化降解作用

采用溶胶-凝胶法制备了掺杂硼的纳米TiO2,并用X射线衍射、红外光谱表征了产物的组成和结构.用制备好的掺硼纳米TiO2作为光催化剂对农药毒死蜱进行了光催化降解研究.结果表明:通过掺硼可以显著提高TiO2对毒死蜱的光催化效率.在硼掺杂量为3%,催化剂煅烧温度为500℃,煅烧时间为1 h,催化剂用量为15%时,毒死蜱的降解率最高.     

Using neural networks to represent potential surfaces as sums of products

By using exponential activation functions with a neural network (NN) method we show that it is possible to fit potentials to a sum-of-products form. The sum-of-products form is desirable because it reduces the cost of doing the quadratures required for quantum dynamics calculations. It also greatly facilitates the use of the multiconfiguration time dependent Hartree method. Unlike potfit product representation algorithm, the new NN approach does not require using a grid of points. It also produces sum-of-products potentials with fewer terms. As the number of dimensions is increased, we expect the advantages of the exponential NN idea to become more significant.     

表面活性剂流动相用于混合固定相薄层色谱

将表面活性剂流动相用于ＧＤＸ与硅胶Ｇ混合固定相薄层色谱,用表面活性剂、乙醇和水的混合流动相对一些水溶性食品染料及指示剂等进行了分离,得到满意的结果,可以部分代替非极性键合相薄板。     

~1H SPIN SYSTEM IDENTIFICATIONS OF S-SULFONATED INSULIN B CHAIN WITH 2D-NMR

Resonance assignments of the ~1H spectrum of insulin are the basis on which to investi-gate its solution conformation by using NMR method. Owing to the complicated aggregationbehaviour of the molecule to give broadened n. m. r. lines, only limited resonance assignmentshave been reported. S-sulfonated A and B chains of insulin gave ~1H spectra with good reso-lutions. Based on the 500 MHz absolute 2D-COSY spectrum and 400 MHz phase sensitiveDQF-COSY, Relayed-COSY and NOESY spectra of B chain recorded in D_2O, all of thespin system identifications of the non-labile protons in the S-sulfonated B chain of insulinwere reported including the specific resonance assignments of eight residues: B_3Asn, B_9Ser,B_(16)Tyr, B_(22)Arg, B_(26)Tyr, B_(27)Thr, B_(28)Pro and B_(29)Lys. The pK values of B_(16) and B_(26) tyrosineare 10.65 and 10.60 respectively from pH titration.     

Li3Al和Li2B2的稳定性和垂直激发态光谱

对于较大的簇合物，电子衍射技术或许能得到有用的信息[1].而小簇合物Mn（3≤n《50）的几何结构就没有标准的方法加以测定．对于二聚体和三聚体[2-5]，振动光谱和转动光谱能够得到精确的核间距和基态的势能面．当n》4时，若要得到一个有价值的结果，需要进行很复杂的振动结构分析．对于这类小簇合物只能得到它们的吸收光谱[6-7].因此，计算其垂直激发态就具有非常重要的意义．用量子化学解释簇合物的吸收光谱已成为一个非常诱人的课题[8].从头算对碱金属簇合物的垂直激发态计算，并与实验光谱进行比较，已成功地预测了一些碱金属簇合物的基…