Generalized extreme learning machine acting on a metric space

Functional data learning is an extension of traditional data learning, that is, learning the data chosen from the Euclidean space ${\mathbb{R}^{n}}$ to a metric space. This paper focuses on the functional data learning with generalized single-hidden layer feedforward neural networks (GSLFNs) acting on some metric spaces. In addition, three learning algorithms, named Hilbert parallel overrelaxation backpropagation (H-PORBP) algorithm, ν-generalized support vector regression (ν-GSVR) and generalized extreme learning machine (G-ELM) are proposed to train the GSLFNs acting on some metric spaces. The experimental results on some metric spaces indicate that GELM with additive/RBF hidden-nodes has a faster learning speed, a better accuracy, and a better stability than HPORBP algorithm and ν-GSVR for training the functional data. The idea of GELM can be used to extend those improved extreme learning machines (ELMs) that act on the Euclidean space ${\mathbb{R}^{n}, }$ such as online sequential ELM, incremental ELM, pruning ELM and so on, to some metric spaces.     

The α- to β-Si3N4 transformation in the presence of liquid silicon

The compacts consisted of , -Si₃N₄ and free silicon are heat treated in the range 1650° C to 1750° C in an argon atmosphere in order to observe the following behaviours; the to phase transformation and variations of the microstructure during heat treatment in silicon nitride. For the microstructural observation of the heat treated specimens, the same grains in the polished surface were investigated before and after eliminating the retained silicon by etching. The to phase transformation, in this case, occurs via silicon melts irrespective of added -Si₃N₄. Both and phases are soluted and precipitated into molten silicon and their morphology are changed from an equiaxed shape to prismatic one. Although elongated grains are precipitated at low temperature or in the early stage of heat treatment, fine precipitated grains are mainly observed with increasing heat treating temperature.     

Dielectric properties and microstructural behaviour of B-site calcium-doped barium titanate ceramics

Dielectric properties and microstructural behaviour of Ba_1–x Sr _x Ti_1–y Ca _y O_3–y ceramics, where strontium and calcium were doped on the barium and titanium sites, respectively, within the range 0x0.24 and 0y0.05, were investigated. Calcium addition decreased the tetragonality,c/a, increased the unit cell volume, and lowered the Curie temperature, which were all attributed to the occupancy of Ca²⁺ ions on titanium sites. When sintered at a low oxygen partial pressure of 10^–9 MPa, a resistivity higher than 10¹¹ cm was maintained for the formulations containing B-site calcium substitution more than 0.5 mol %. With increasing the amount of calcium addition, the Curie peak was depressed and completely broadened for the compositions with calcium addition more than 3 mol %, where the average grain size was smaller than 1 m. Co-firing with nickel electrodes in a reducing atmosphere also depressed the Curie peak and inhibited the grain growth due to the diffusion of nickel into the dielectrics.     

利用CFD技术研究四边形通道内的流场

了解认知隧道内的流体力学的特点对于隧道火灾的研究和隧道安全建设有着非常大的意义。为了对隧道火灾的特点进行研究,我们使用商用CFD程序STAR-CD对速度场和压力分布进行了数值模拟。模拟过程中在同一入口位置的情况下改变四个出口位置,出口位置与模具高度的数值对别分别为L/H=0,0.375,0.75,1.125。同时还介绍了对于火灾救援和人员疏散的有效措施。     

The diffusion and solubility of oxygen in solid nickel

Solid-state electrochemical measurements using various experimental procedures were made with the double cell: $$ Ni + NiO|ZrO_2 + Y_2 O_3 |Ni + \underline O |ZrO_2 + Y_2 O_3 |Ni + NiO $$ to determine the diffusivity and thermodynamic functions of oxygen dissolved in solid nickel. Non-steady state diffusion of oxygen in the specimen was caused by applying a preselected potential between the reference and specimen electrodes and was monitored by measuring time-dependent potentials and/or currents. The following results were obtained for the diffusivity of oxygen and the solubility of oxygen in nickel in equilibrium with NiO: $$D{\text{ = 4}}{\text{.9 }} \times {\text{ 10}}^{{\text{ - 2}}} {\text{ exp}}\left( {{\text{ - }}\frac{{{\text{164 kJ/mole}}}}{{{\text{R}}T}}} \right){\text{cm}}^{\text{2}} /{\text{sec (850 to 1400 }}{}^{\text{o}}{\text{C)}}$$ $$C_{\text{O}}^s {\text{ = 8}}{\text{.3 exp}}\left( { - \frac{{55{\text{kJ/mole}}}}{{{\text{R}}T}}} \right){\text{at}}{\text{. pct (800 to 1000 }}{}^{\text{o}}{\text{C)}}$$ The thermodynamic and transport behaviors of oxygen in solid nickel were fairly well described by a simple quasi-regular model and an interstitial diffusion model, respectively.     

Diffusion of additives and plasticizers in poly(vinyl chloride): 4. A programmed temperature technique for the determination of the diffusion parameters of three dialkylphthalates

A programmed temperature technique has been developed for the measurement of the self-diffusion coefficients of additives in polymers. This has been applied to the diffusion of three dialkylphthalate plasticizers in poly(vinyl chloride). Values of the self-diffusion coefficients and the calculated activation energies obtained in this way have been compared with values previously obtained from measurements at constant temperature. It has been shown that the programmed temperature technique gives a very rapid method for the estimation of the activation energy of self-diffusion.     

Dispersion of flexible polymer chains in confined geometries

Molecular model approach has been used to predict the dispersion characteristics of flexible polymer chains in confined geometries. The analysis ranges from the early stage dispersion to the steady Taylor dispersion of the simple linear dumbbell model polymer chains. For the early stasje dispersion the ray method was applied; an Aris type moments rnothod was useful for the Taylor dispersion. Two parallel plates were chosen as a confining geometry and the specific initial condition of a point source in the midway of the gap was chosen for simplicity. It was found that the qualitative difference in dispersion properties of deformable polymer chains starts from the early stage compared with those of single Brownian particles. And it turns out that one parameter, which is similar to the relative spacing of the dumbbell to the gap of confining geometries, is useful to see the dispersion characteristics of the dumbbells.     

Catalytic Combustion of Methane over Rare Earth Stannate Pyrochlore

Well-defined Ln₂Sn₂O₇ powders (Ln = La, Sm and Gd) with a phase-pure pyrochlore structure were synthesized by hydrothermal reaction. The catalytic activities of Ln₂Sn₂O₇ powders for methane combustion were measured. Methane oxidation started at 500 °C and increased with oxidation temperature. Catalytic methane combustion is strongly influenced by the presence of oxygen vacancies that form by breaking Sn–O lattice bonds as the temperature increases. Addition of manganese to the rare earth pyrochlores improved methane oxidation activity. Manganese-doped samarium stannate pyrochlore (Sm₂Sn_1.8Mn_0.2O₇) shows highest the catalytic activity. Light-off and complete oxidation temperatures were measured at about 400 and 650 °C, respectively.     

Analysis of the correlation between acoustic noise and vibration generated by a multi-layer ceramic capacitor

Acoustic noise generated by a multi-layer ceramic capacitor (MLCC) makes users uncomfortable, so the problem must be analyzed to reduce the noise. There is a correlation between the acoustic noise and the vibration of MLCCs and the circuit board. Therefore, the acoustic noise problem must be investigated from a vibration perspective. In this study, the acoustic noise-generating mechanism was investigated, and the relationship between the characteristics of the noise and the dynamic characteristics of the circuit board with MLCC was analyzed. And a correlation criterion was proposed to predict the acoustic noise using the vibration response of the circuit board.     

An experimental investigation of mixing of carbon nanotube/polymer composite in a batch-type screw mixer

We develop a miniaturized batch-type screw mixer (BSM) for uniform mixing of polymer resin and nanoparticles, based on the stretching of material elements. This stretching is induced by the combination of recirculating cross-sectional flows in deep channels of the screw and high shear stress developed at flight regions. The BSM is used to produce a polymer nano-composite composed of multi-walled carbon nanotubes and polydimethylsiloxane resin. The mixing performance of the BSM is characterized quantitatively by estimating two different types of mixing efficiencies (i.e., dispersive mixing and distributive mixing) via transmitted light microscope images. The developed BSM highly improves the mixing performance rather than that of a conventional ultrasonic mixing device.