Polyimide/montmorillonite nanocomposites based on thermally stable, rigid-rod aromatic amine modifiers

The preparation and processing of most of polymer/clay nanocomposites need high temperature. This limited the application of commonly used organic modifiers of long carbon-chain alkyl ammonium salts because of their low thermal stability. In this study, we synthesized two novel thermally stable, rigid-rod aromatic amines. Montmorillonite (MMT) treated by these amines exhibited larger layer-to-layer spacing, higher thermal stability than that treated by commonly used 1-hexadecylamine and also high ion-exchange ratio (>95%). They were applied to prepare nanocomposites with polyimide (PI) by in situ polymerization. XRD, TEM were used to obtain the information on morphological structure of PI/MMT nanocomposites. DMA, TGA, DSC, universal tester were applied to characterize the mechanical and thermal properties of the nanocomposites. When the MMT content was below 3 wt%, the PI/MMT nanocomposites were strengthened and toughened at the same time. The introduction of a small amount of MMT also led to improvement in thermal stability, slight increase in glass transition temperature, marked decrease in coefficient of thermal expansion and decrease in solvent uptake. MMT treated by these aromatic amines exhibited better dispersibility and (probably) interfacial interaction with PI matrix than that treated by 1-hexadecylamine. The nanocomposites based on these MMT resultantly exhibited better mechanical, thermal and solvent resistance properties than those based on 1-hexadecylamine treated MMT.     

自适应序列生成的建筑能耗预测

提出一种基于强化学习的生成对抗网络(Reinforcement learning-based Generative Adversarial Networks,Re-GAN)能耗预测方法.该算法将强化学习与生成对抗网络相结合,将GAN(Generative Adversarial Nets)中的生成器以及判别器分别构建为强化学习中Agent(生成器)以及奖赏函数.在训练过程中,将当前的真实能耗序列作为Agent的输入状态,构建一组固定长度的生成序列,结合判别器及蒙特卡洛搜索方法进一步构建当前序列的奖赏函数,并以此作为真实样本序列后续第一个能耗值的奖赏.在此基础之上,构建关于奖赏的目标函数,并求解最优参数.最后使用所提算法对唐宁街综合大楼公开的建筑能耗数据进行预测试验,实验结果表明,所提算法比多层感知机、门控循环神经网络和卷积神经网络具有更高的预测精度.     

基于轮廓几何稀疏表示的刚性目标模型及其分级检测算法

刚性目标轮廓具有明显几何特性且不易受光照、纹理和颜色等因素影响.结合上述特性和图像稀疏表示原理,提出一种适用于刚性目标的分级检测算法.在基于部件模型(Part-based model, PBM)的框架下,采用匹配追踪算法将目标轮廓自适应地稀疏表示为几何部件的组合,根据部件与目标轮廓的匹配度,构建描述部件空间关系的有序链式结构.利用该链式结构的有序特性逐级缩小待检测范围,以匹配度为权值对各级部件显著图进行加权融合生成目标显著图. PASCAL图像库上的检测结果表明,该检测方法对具有显著轮廓特征的刚性目标有较好的检测结果,检测时耗较现有算法减少约60%~90%.     

基于铌酸锂薄膜的Y波导集成调制器设计

近年来,基于薄膜铌酸锂（TFLN）平台的高性能电光调制器以其小体积、低能量损耗等特点受到广泛关注。本文提出了一种新的具有垂直电极结构的Y波导铌酸锂薄膜电光调制器。研究了调制器的半波电压与缓冲层厚度之间的关系,优化了Y波导的设计参数,最后设计得到插入损耗<5 dB, 半波电压<1.5 V的高性能调制器。本文不仅为基于TFLN平台的小型化波导的设计和实现提供了思路,而且为制造高性能和多功能的电光调制器提供了实验依据。     

凝胶注模工艺制备高强度多孔氮化硅陶瓷

采用凝胶注模成型工艺,成功地制备了具有高强度、结构比较均匀并有较高气孔率的氮化硅多孔陶瓷。本文研究了制得的多孔氮化硅的力学性能和微观结构,并讨论了获得高性能的原因。结果表明,采用适当的成型条件可制备出结构均匀、强度高、加工性能优良的坯体,烧成的多孔氮化硅陶瓷强度均>150MPa,气孔率>50％。SEM照片显示气孔是由长柱状β-Si3N4晶搭接而成的。均匀的气孔分布和柱晶结构是获得高性能的主要原因。     

Ebola Entry Inhibitors Discovered from Maesa perlarius

Ebola virus disease (EVD), a disease caused by infection with Ebola virus (EBOV), is characterized by hemorrhagic fever and a high case fatality rate. With limited options for the treatment of EVD, anti-Ebola viral therapeutics need to be urgently developed. In this study, over 500 extracts of medicinal plants collected in the Lingnan region were tested against infection with Ebola-virus-pseudotyped particles (EBOVpp), leading to the discovery of Maesa perlarius as an anti-EBOV plant lead. The methanol extract (MPBE) of the stems of this plant showed an inhibitory effect against EBOVpp, with an IC₅₀ value of 0.52 µg/mL, which was confirmed by testing the extract against infectious EBOV in a biosafety level 4 laboratory. The bioassay-guided fractionation of MPBE resulted in three proanthocyanidins (procyanidin B2 (1), procyanidin C1 (2), and epicatechin-(4β→8)-epicatechin-(4β→8)-epicatechin-(4β→8)-epicatechin (3)), along with two flavan-3-ols ((+)-catechin (4) and (−)-epicatechin (5)). The IC₅₀ values of the compounds against pseudovirion-bearing EBOV-GP ranged from 0.83 to 36.0 µM, with 1 as the most potent inhibitor. The anti-EBOV activities of five synthetic derivatives together with six commercially available analogues, including EGCG ((−)-epigallocatechin-3-O-gallate (8)), were further investigated. Molecular docking analysis and binding affinity measurement suggested the EBOV glycoprotein could be a potential molecular target for 1 and its related compounds.     

基于分类DQN的建筑能耗预测

本文提出一种可用于建筑能耗预测的基于KNN分类器的DQN算法——K-DQN.其在利用马尔科夫决策过程对建筑能耗进行建模时,针对大规模动作空间问题,将原始动作空间缩减进而提高算法的预测精度及收敛速率.首先, K-DQN将原始动作空间平均划分为多个子动作空间,并将每个子动作空间对应的状态分为一类,以此构建KNN分类器.其次,利用KNN分类器,将不同类别相同次序动作进行统一表示,以实现动作空间的缩减.最后,K-DQN将状态类别概率与原始状态相结合,在构建新状态的同时,帮助确定缩减动作空间内每一动作的具体含义,从而确保算法的收敛性.实验结果表明,文章提出的K-DQN算法可以获得优于DDPG、DQN算法的能耗预测精度,且降低了网络训练时间.     

Research on stability of a slope due to underground mining

This paper will present a detailed analysis of the deformation mechanism and stability assessment of the slope through field investigations, numerical modeling and measurements. Field investigation indicated that three thin coal seams encountered large mined-out area at one side and free surface of hill slope at the other side, which lead to the caving of roof strata movement, ground movement and crown crack along the preferred orientations of joints. The three-dimensional numerical modeling study on the case demonstrated that the plasticity failure occurred gradually along with the extension of mined-out area in depth. When the depth of mining reached the verge defined by the seismic prospecting method, a large mount of tension failure occurred on the crown of the slope. The factor of safety was 1.36 calculated by the shear strength reduction technique, which indicated the slope was in stable state. The measurement showed that the residual deformation occurred before 1998 and became stable subsequently, which indicated that the residual deformation almost finished and the slope is in stable state.     

四川蘑菇山矿区控矿构造特征及构造控矿模式

层间破碎带和断裂构造是蘑菇山矿区最为重要的控矿因素,直接控制着矿床(体)的产出及空间分布.不同级别、不同性质及不同规模的断裂构造对矿床的控制具有明显的差异.通过控矿断裂构造特征研究进而揭示矿区构造控矿规律,对矿床深部找矿前景评价有重要意义.     

Surface decomposition and annealing behavior of GaN implanted with Eu

Investigations on surface decomposition of GaN implanted with low energy (80keV) Eu ion to a low dose (1×1014cm-2), and its annealing behavior under high temperature (1050℃) in N2 are performed. The as-grown, as-implanted and annealed GaN films are characterized by proton elastic scattering (PES), Rutherford backscattering spectrometry (RBS), photoluminescence (PL) and atomic force microscopy (AFM). The results show that Eu ion implantation induces radiation defects and decomposition of GaN. The GaN surface decomposition is more serious during high temperature annealing. The atomic ratio of N in as-grown, as-implanted and annealed GaN film is 47 at.%, 44 at.% and 40 at.%, respectively. As a result, a rough Ga-rich layer is formed at the surface, though the lattice defects are partly removed after high temperature annealing.