Preparation of V doped lanthanum silicate electrolyte ceramics by combustion method and study on conductance mechanism

Vanadium doped La_9.33Si_6−xV_xO_26+0.5x (x = 0.5, 1.0, 1.5) (LSVO) electrolyte powder was prepared by combustion method at 600°C for 5-7 min. The powder was sintered at 1500°C for 3 hours to prepare LSVO ceramics. XPS, IR, XRD, and EIS analysis show that V⁵⁺ doping replaces Si⁴⁺ in [SiO₄] to form [Si(V)O₄] tetrahedron. With the increase in x, the lattice volume increase. When x = 2.0, the LaVO₄ phase was formed, indicating that the limit doping amount of V⁵⁺ replacing Si⁴⁺ is x ≤ 1.5. The conductivity of LSVO increases significantly with the increase in x (x ≤ 1.0), which attributed to the defect reaction caused by V⁵⁺ doping. The addition of the interstitial oxygen Oi* in 6₃ channels and the increase of lattice volume leads to increased conductivity. When x = 1.0, the highest conductivity is 1.46 × 10⁻² S·cm⁻¹ (800°C). The doping enhancement conductivity mechanism is the Interstitial oxygen defect-Lattice volume composite enhancement mechanism.     

Safety evaluation and the static-dynamic coupling analysis of counter-arched slab in plunge pool

The stability of plunge pool slab of drainage structure is crucial to energy dissipation, but the working conditions of counter-arched slab are very complex and its stress procedure is considerably nonlinear and coupling. Therefore, a nonlinear static and dynamic coupling method to analyze the counter-arched slab configuration is put forward, which can reflect the coupling of dynamic loads, slab, anchor bars, groundwork and abutment as well as the dynamic procedure of instability. Various nonlinear factors and static-dynamic coupling are taken into consideration, and the working conditions, stress mechanism, dynamic instability procedure and influential instability factors are revealed. The proposed method thus provides a comprehensive safety evaluation method of plunge pool, which further provides an important theoretical basis for engineering design of counter-arched slab in plunge pool. Supported by the National Science Fund for Distinguished Young Scholars (Grant No. 50725929) and the National Natural Science Foundation of China (Grant Nos. 50539060, 50679052, 50779044)     

Flux pinning effect of cubic equiaxed morphology and its Ti stabilizing in Nb<Subscript>3</Subscript>Sn superconductors

Two sets of internal-Sn Nb3Sn superconducting strands were fabricated through RRP method, one with 2 wt% of Ti alloyed in Sn core and the other just pure Sn. Four reaction temperatures of 650℃, 675℃, 700℃ and 725℃ and 128 h duration were applied for A15 phase formation heat treatment after Cu-Sn alloying procedure of 210℃/50 h + 340℃/25 h. For the heat-treated coil samples, transport non-Cu JC was examined through standard 4-probe technique and phase microstructure was observed by means of Field Emission Sc...     

Characterization of catalysts and membrane in DMFC lifetime testing

The lifetime and performance of a direct methanol fuel cell (DMFC) were investigated to understand the correlation between the structure of catalysts/membrane and cell performance versus time. The cell polarization and performance curves were obtained during the DMFC operation with the time. The catalysts and Nafion^® membrane of the membrane electrode assembly (MEA) from the lifetime test were comprehensively examined by XRD, HRTEM, FTIR and Raman spectroscopy techniques. The results revealed that there was significant performance degradation during the first 200 h operation; while the degradation was slowing down between 200 and 704 h operation. The degradation became worse after 1002 h operation. The increases of the catalyst particle size from both anode and cathode catalysts were observed after the DMFC lifetime test. The changes of microstructure, surface composition, the interfacial structure of the MEA, and the aging of Nafion^® under the DMFC lifetime tests were also observed.     

微信流量特性分析与建模

微信是现代互联网的主要应用之一,到目前为止有关微信流量特性分析与建模的研究较少.本文以微信流量为研究对象,分析验证微信流量同时具有自相似性和突发性.针对这两种特性进行微信流量建模,采用线性分形稳定噪声模型刻画微信流量特性,完成了模型的参数估算和效果分析.本文的研究成果是后续的网络性能分析、网络流量监管等的基础.     

一种基于深度学习的视网膜病变图像识别方法

为解决医学上糖尿病性视网膜病变图像人工识别困难、精度差等问题,提出一种基于多特征融合的卷积神经网络识别方法。在VGG-16模型的基础上,通过融合每层网络上的局部特征,增强模型的特征提取能力。选用Softmax分类器,使病变图像识别更加准确。使用OpenCV图像处理工具采用加噪、上下左右不同角度翻转、调节对比度等5种方式扩充训练集。实验结果表明,基于多特征融合的深度学习框架图像识别系统在数据集上的平均识别精度达到94.23%,相较于Alex-Net、Google-Net、Compact-Net、ResNet-101等模型分别提高了10.56%、7.80%、6.01%、0.02%,验证了该方法的有效性。该模型具有很好的鲁棒性。     

基于Siamese LSTM的中文多文档自动文摘模型

在文本信息数量迅速增长的环境下,为提升阅读效率,提出一种基于深度学习的多文档自动文本摘要模型。在传统文摘模型的基础上将Siamese LSTM深度学习网络应用到文本相似度计算中,计算曼哈顿距离来表征文本相似度,并采用去除停用词的方法改进该网络模型以提升计算效率。实验结果表明,使用Siamese LSTM与传统余弦相似度等方法相比,生成的文摘在语义方面更贴近主题,质量更高,整个文摘系统的工作效率也显著提升。     

动态多元分类属性的Hyperbolic Tree可视化

以表格和链接节点树为形式的分类数据展示,存在层次关系不明显和无法表达大规模层次结构的局限性。针对表格形式存在的分类型数据,本文设计和实现了基于Hyperbolic Tree(双曲树)的动态多元分类属性可视化方案。该方案首先根据给定的多元分类属性序进行基于表格数据的分支结构树构建,进而采用双曲线树对分支结构树进行可视化,最后通过调整多元分类属性的序实现动态可视化。同时,以国家电网营业网点分布数据为对象,对更改主次分类属性的动态可视化进行实验,既可以直观地展示网点的地理区域分布特性,也可用于辅助网点的分类与聚类。     

柴油机自动减压起动装置的有限元分析

分析了柴油机自动减压装置的工作原理,针对在潍柴WP10型柴油机上加装的减压装置,利用Solidworks Simulation对压板进行了静力分析,找出了应力最大的部位和最大值,并按照Von Mises准则,判别压板满足刚度要求。     

宁夏固原市主要农作物生产水足迹分析

水足迹理论为解决水资源短缺问题提供了新的研究方向。基于水足迹理论,借助CROPWAT8. 0软件,利用《宁夏统计年鉴》和《宁夏水资源公报》等资料,计算分析了宁夏固原市2009-2017年主要作物的蓝水、绿水和灰水足迹。结果表明:蓝水足迹占比最高,绿水足迹次之,灰水足迹最少。表明固原市农作物生产用水以蓝水为主,绿水为辅,灰水足迹虽然占比较少但逐年递增。建议当地根据水资源具体情况,调整作物种植模式,提高绿水利用率,减少氮肥使用率,使有限的水资源获得更大的效益,从而满足粮食安全和水生态安全的需求。