用人工神经网络处理谷物成分分析

本文用人工神经网络处理谷物的付立叶变换近红外漫反射光谱,对谷物中含量在10~(-1)～10~(-3)的蛋白质、脂肪和6种人体必需氨基酸定量分析数据进行了解析,分析结果与经典化学方法没有系统偏差,且优于逐步回归分析法的结果。     

Establishing structure-property correlations and classification of base oils using statistical techniques and artificial neural networks

The present paper describes various classification techniques like cluster analysis, principal component (PC)/factor analysis to classify different types of base stocks. The API classification of base oils (Group I-III) has been compared to a more detailed NMR derived chemical compositional and molecular structural parameters based classification in order to point out the similarities of the base oils in the same group and the differences between the oils placed in different groups. The detailed compositional parameters have been generated using and nuclear magnetic resonance (NMR) spectroscopic methods. Further, oxidation stability, measured in terms of rotating bomb oxidation test (RBOT) life, of non-conventional base stocks and their blends with conventional base stocks, has been quantitatively correlated with their NMR and elemental (sulphur and nitrogen) data with the help of multiple linear regression (MLR) and artificial neural networks (ANN) techniques. The MLR based model developed using NMR and elemental data showed a high correlation between the ‘measured’ and ‘estimated’ RBOT values for both training (R=0.859) and validation (R=0.880) data sets. The ANN based model, developed using fewer number of input variables (only NMR data) also showed high correlation between the ‘measured’ and ‘estimated’ RBOT values for training (R=0.881), validation (R=0.860) and test (R=0.955) data sets.     

Irreversible deformation of macromolecular networks

Based upon a thermodynamical approach, the generalized Onsager type of relaxation of van der Waals networks is presented. By linearly and identically coupling the set hidden variables to the network, the memory function of the system can be related to the equilibrium strain-energy function. The relaxation behavior of real networks on stretching can quantitatively be described by means of a distribution of relaxation times known from small strain experiments. Some new and interesting conclusions are discussed as to how the macroscopically non-linear visco-elastic response might be interpreted.     

Simulation of aerated lagoon using artificial neural networks and multivariate regression techniques

The aim of this study was to develop an empirical model that provides accurate predictions of the biochemical oxygen demand of the output stream from the aerated lagoon at International Paper of Brazil, one of the major pulp and paper plants in Brazil. Predictive models were calculated from functional link neural networks (FLNNs), multiple linear regression, principal components regression, and partial least-squares regression (PLSR). Improvement in FLNN modeling capability was observed when the data were preprocessed using the PLSR technique. PLSR also proved to be a powerful linear regression technique for this problem, which presents operational data limitations.     

π-Bindungsbeitrag in Boranen, 2. Mitt.:Ab initio- und Kraftfeld-Rechnungen von Vinyldifluorboran

Ab initio calculations on vinyldifluoroborane yield a -contribution of 23 kJ/mol (5.5 kcal/mol) mainly due to the C=C bond and not to the lone pairs of the fluorine atoms. The rotational barrier was also determined. The force field calculations favour the interpretation of the bond structure derived from the theab initio results.

     

Force Constants and Bond Orders of Nitrogen Bonds

The force constants and the corresponding bond orders of nitrogen bonds have been calculated from the vibrational spectra (infrared and Raman spectra) of a great number of nitrogen compounds. Plotting the maximum bond order of stable nitrogen bonds against the sum of Pauling's electronegativities of the bonding partners (Σx) leads to one continuous curve for the N? X bonds where X represents elements of the first and the second short period of the periodic table. Furthermore, when the bonds formed between these elements are arranged in a coordinate system in such a way that the position of each bond is determined by the difference between the electronegativities of the bonding partners (Δx along the ordinate) and the sum of the electronegativities of the bonding partners (Σx along the abscissa), the bonding partners capable of forming multiple bonds all lie within a closed domain, where their position can be correlated with their polymerizability and other reactivities of the multiple bonds. Also discussed are the orders of bonds between nitrogen and some transition elements. In an appendix, the present methods used to calculate force constants and bond orders are surveyed.     

Interpenetrating polymer networks of poly(dimethyl siloxane–urethane) and poly(methyl methacrylate)

Simultaneous IPNs of poly(dimethyl siloxane-urethane) (PDMSU)/poly(methyl methacrylate) (PMMA) and related isomers have been prepared by using new oligomers of bis(β-hydroxyethoxymethyl)poly(dimethyl siloxane)s (PDMS diols) and new crosslinkers biuret triisocyanate (BTI) and tris(β-hydroxylethoxymethyl dimethylsiloxy) phenylsilane (Si-triol). Their phase morphology have been characterized by DSC and SEM. The SEM phase domain size is decreased by increasing crosslink density of the PDMSU network. A single phase IPN of PDMSU/PMMA can be made at an M_c = 1000 and 80 wt % of PDMSU. All of the pseudo- or semi-IPNs and blends of PDMSU and PMMA were phase separated with phase domain sizes ranging from 0.2 to several micrometers. The full IPNs of PDMSU/PMMA have better thermal resistance compared to the blends of linear PDMSU and linear PMMA. © 1993 John Wiley & Sons, Inc.     

Retention modeling and simultaneous optimization of pH value and gradient steepness in RP‐HPLC using feed‐forward neural networks

A novel approach is proposed for the simultaneous optimization of mobile phase pH and gradient steepness in RP‐HPLC using artificial neural networks. By presetting the initial and final concentration of the organic solvent, a limited number of experiments with different gradient time and pH value of mobile phase are arranged in the two‐dimensional space of mobile phase parameters. The retention behavior of each solute is modeled using an individual artificial neural network. An “early stopping” strategy is adopted to ensure the predicting capability of neural networks. The trained neural networks can be used to predict the retention time of solutes under arbitrary mobile phase conditions in the optimization region. Finally, the optimal separation conditions can be found according to a global resolution function. The effectiveness of this method is validated by optimization of separation conditions for amino acids derivatised by a new fluorescent reagent.     

Graft copolymer of polymethyl methacrylate-g-polycarbonate urethane and the fully gelled grafted networks

We have synthesized and characterized the graft copolymer of polymethyl methacrylate (PMMA)-graft-polycarbonate urethane (PCU) (designated PMMA-g-PCU) and the fully gelled mutually crosslinked network composed of both PMMA and PCU (designated as PMMA-PCU-N). Unlike the previously synthesized one phase PCU/PMMA simultaneous interpenetrating polymer networks (IPN's), both of these new materials exhibited two distinct T_g's (from DSC and DMA studies) at all compositions which we have studied. The two-phase nature of the graft and network polymer was confirmed by TEM with domain sizes smaller than those found in the blends of the linear PCU and PMMA chains. In comparison with the corresponding IPN's, the PMMA-g-PCU and the PMMA-PCU-N were thermally less stable and had a lower ultimate strength. © 1992 John Wiley & Sons, Inc.