Cuso4‐amine redox‐initiated radical polymerization of methyl methacrylate mediated by a CuCl2 complex: Homogeneous reverse ATRP

Kinetic results of CuSO₄/2,2'‐bipyridine(bPy)‐amine redox initiated radical polymerization of methyl methacrylate (MMA) at 70 to 90 °C in dimethylsulfoxide suggest that such initiation is characteristic of a slow rate and a low initiator efficiency, but tertiary amines exhibit a relatively higher rate. UV‐Vis spectroscopy confirms the alpha‐amino functionality of PMMA chains. CuCl₂/bPy successfully mediates the redox‐initiated radical polymerization of MMA with aliphatic tertiary amines in a fashion of slow‐initiated reverse atom transfer radical polymerization (ATRP), i.e. both the initiator efficiency of aliphatic tertiary amines and the average molecular weight of PMMA increase gradually, while the molecular weight distribution remains narrow but become broader with the conversions. As the PMMA chains contain alpha amino and omega C‐Cl moieties, UV‐induced benzophenone‐initiated radical polymerization and Cu^ICl/bPy‐catalyzed ATRP initiated from PMMA lead to block copolymers from terminal functionalities. © 2014 Wiley Periodicals, Inc. J. Polym. Sci., Part A: Polym. Chem. 2014 , 52, 2562‐2578     

基于炼焦煤族组成和结构参数的焦炭质量预测模型及其成焦机理

以五种炼焦煤和44组配合煤为研究对象,在40 kg小焦炉环境下完成煤杯炼焦实验,以煤全组分分离所获得的煤重质组、密中质组和疏中质组收率Y_(HC)、Y_(DMC)、Y_(LMC)及反映煤中氢键缔合情况和脂肪链长短或支链化程度的红外光谱参数I3、I4为主要指标,通过BP神经网络分析方法建立了焦炭质量预测模型,并讨论了模型的特点,分析了新模型下的成焦机理。结果表明,使用新的煤组成结构参数预测焦炭质量具有一定优势,成焦率(CR)、显微强度(MSI)、粒焦反应性(PRI)和反应后强度(PSR)的预测值和实测值有较好一致性,对y=x的拟合相关系数分别达到0.986、0.982、0.956和0.926。模型对CR、MSI和PRI的预测效果较好,九个预测样本的平均偏差分别为0.53%、1.58%和1.28%;但对反应后强度PSR预测效果较差,平均偏差在12.22%。研究结果为建立炼焦配煤新方法提供了良好基础。     

The H2+ molecular ion: Low-lying states

Matching for a wavefunction the WKB expansion at large distances and Taylor expansion at small distances leads to a compact, few-parametric uniform approximation found in Turbiner and Olivares-Pilon (2011). The ten low-lying eigenstates of H₂⁺ of the quantum numbers (n,m,Λ,±)$(n,m,\Lambda ,\pm )$  with n=m=0$n=m=0$ at Λ=0,1,2$\Lambda =0,1,2$, with n=1$n=1$, m=0$m=0$ and n=0$n=0$, m=1$m=1$ at Λ=0$\Lambda =0$ of both parities are explored for all interproton distances R$R$. For all these states this approximation provides the relative accuracy ?10⁻⁵$?10^{-5}$ (not less than 5 s.d.) locally, for any real coordinate x$x$ in eigenfunctions, when for total energy E(R)$E\left(R\right)$ it gives 10-11 s.d. for R∈[0,50]$R\in [0,50]$  a.u. Corrections to the approximation are evaluated in the specially-designed, convergent perturbation theory. Separation constants are found with not less than 8 s.d. The oscillator strength for the electric dipole transitions E1$E1$ is calculated with not less than 6 s.d. A dramatic dip in the E1$E1$ oscillator strength f_{1sσg−3pσu}$f_{1s{\sigma }_g-3p{\sigma }_u}$ at R∼R_eq$R\sim R_{eq}$ is observed. The magnetic dipole and electric quadrupole transitions are calculated for the first time with not less than 6 s.d. in oscillator strength. For two lowest states (0,0,0,±)$(0,0,0,\pm )$ (or, equivalently, 1sσ_g$1s{\sigma }_g$ and 2pσ_u$2p{\sigma }_u$ states) the potential curves are checked and confirmed in the Lagrange mesh method within 12 s.d. Based on them the Energy Gap between 1sσ_g$1s{\sigma }_g$ and 2pσ_u$2p{\sigma }_u$ potential curves is approximated with modified Pade Re^−R[Pade(8/7)](R)$R{e}^{-R}\left[Pade(8/7)\right]\left(R\right)$ with not less than 4-5 figures at R∈[0,40]$R\in [0,40]$ a.u. Sum of potential curves E_1sσg+E_2pσu$E_{1s{\sigma }_g}+E_{2p{\sigma }_u}$ is approximated by Pade 1/R[Pade(5/8)](R)$1/R\left[Pade(5/8)\right]\left(R\right)$ in R∈[0,40]$R\in [0,40]$ a.u. with not less than 3-4 figures.     

On The diophantine equation Fn+Fm=2a

In this paper, we find all the solutions of the title Diophantine equation in positive integer variables (n, m, a), where F_k is the kth term of the Fibonacci sequence. The proof of our main theorem uses lower bounds for linear forms in logarithms (Baker's theory) and a version of the Baker-Davenport reduction method in diophantine approximation.     

基于灰关联分析的GA-BP神经网络在汽车油耗估算中的应用

从设计参数特征入手分析影响汽车油耗的因素,利用灰关联分析方法,解析了各设计参数对汽车油耗的影响程度,选择其中灰关联度较大的设计参数作为输入数据,综合工况油耗作为输出数据,构建6-5-1层结构的BP神经网络预测模型,并利用遗传算法获得优化后的BP神经网络的权值和阈值,然后训练BP神经网络得到最优值,最后以国内市场340款汽车作为研究样本,进行有效性验证.研究结果表明,模型利用灰关联分析获得影响汽车油耗的主要因素,简化了网络结构;与优化前的BP神经网络相比,具有更高的预测精度和可靠性.     

基于遗传算法和神经网络的注塑工艺参数优化

基于Moldflow模拟仿真的结果,结合GA算法优化BP网络的结构,建立了模具温度,熔体温度,保压压力,注射速度等工艺参数与塑件体积收缩率的BP网络模型.获得了最优的工艺参数组合,同时预测结果与实际结果吻合.通过神经网络算法(BP)预测注塑工艺参数对塑件质量的影响,可以有效降低其他建模方法的难度和工作量,方法可以推广到塑件其他质量预测过程中.     

水面舰队建立最佳防空编队的数学模型

防空反导是水面舰艇编队最重要的任务之一,我国南海海域辽阔,水面舰艇执行外围岛礁附近海域巡航任务时,往往超出了空中掩护的作战半径,需要自身的对空防御,这时,水面舰艇的编队阵型至关重要.同时,空中目标意图识别是战场态势分析的一个重要部分.以我海军在南海某开阔海域巡逻的水面舰艇编队为例,探究了最佳编队队形的数学模型,并根据所提供的战场空中目标信息,判断目标可能的意图,为威胁判断、火力分配和抗击来袭目标奠定基础.     

基于神经网络和HVS的空域数字水印算法

为了弥补空间域水印算法在提取水印时需要原宿主图像、鲁棒性较差等缺点,提出一种基于BP神经网络和人眼视觉特性(HVS)的空域数字水印算法。先选出原宿主图像对比度函数值最大的前块用于二值水印图像的嵌入,利用神经网络能够逼近任意非线性关系的特点和其自适应性,构建水印信号嵌入前后图像块像素值间的映射关系,实现了水印的盲提取,并将该算法与现有算法进行了对比。实验结果表明,对于JPEG压缩、加噪、剪切和旋转等常见的图像处理攻击,算法具有良好的鲁棒性和不可见性。     

Synthesis of lariat organochalcogenoethers based on azacalix[3]arenes for the potentiometric detection of [UO2] ions

The syntheses of organochalcogen-supported azacalix[3]arenes are described in a one-pot manner in satisfactory yields. A remarkably selective potentiometric response was accomplished for uranyl ions over a variety of other metal ions, including alkali (Na⁺, K⁺), alkaline-earth (Mg²⁺, Ca²⁺, Ba²⁺), transition and heavy metal ions (Co²⁺, Ni²⁺, Cu²⁺, Ag⁺, Fe³⁺, Zn²⁺, Cd²⁺ and Pb²⁺) using an ion-selective electrode based on compound 3 incorporated into a polymeric (PVC) membrane.     

A unique decomposition result for HT factors with torsion free core

We prove that II₁ factors M have a unique (up to unitary conjugacy) cross-product type decomposition around “core subfactors” N⊂M satisfying the property HT of [S. Popa, On a class of type II₁ factors with Betti numbers invariants, Ann. of Math. (2) 163 (2006) 809-899] and a certain “torsion freeness” condition. In particular, this shows that isomorphism of factors of the form Lαi(Z²)?Fni, i=1,2, for Fni⊂SL(2,Z) free groups of rank ni and αj=e2πitj, tj∉Q, implies n₁=n₂.