固体力学中快速多极边界元法研究进展

和快速多极方法相结合，使边界元法处理大规模工程与科学问题变得十分有效,首先概括介绍了快速多极边界元法，接着介绍其精度和效率的验证以及与常规边界元法的比较，并给出在微机机群上的快速多极边界元并行算法,给出了快速多极边界元法的一些应用，其中包括：复合材料的二维、三维模拟，含大量裂纹的二维弹性固体及其疲劳裂纹扩展的模拟．此外还介绍了用于弹塑性问题的快速多极边界元新方法．     

Responses to the H-2Kba mutant involve recognition of syngeneic Ia molecules

Conventional antigens appear to be recognized by T lymphocytes only when associated with major histocompatibility complex (MHC) antigens. Using antigen-specific proliferation as a model for helper T lymphocytes, it has been demonstrated that Ly1+T cells recognize antigen presented in association with syngeneic Ia molecules. In contrast to responses to conventional antigens, however, a large number of studies have suggested that the stimulation of alloreactive Ly1+T cells, and helper T cells specific for allogeneic cytotoxic T lymphocyte (CTL) responses, involve the direct recognition of Ia alloantigens. For the generation of optimal allogeneic CTL activity it has been proposed that Ly1+T cells recognize allo-Ia antigens directly and provide help to pre-CTLs that respond to allo-H-2K and/or D determinants. Thus, the B6.C.H-2bm1 mutant (bm1, formerly referred to as Hz1), which is believed to consist of a substitution of two amino acids in the H-2Kb antigen, has presented a paradox, for it can stimulate strong mixed lymphocyte culture (MLC), graft versus host and CTL responses by T cells of H-2b haplotype mice in the apparent absence of any alloantigenic differences in the I region. We now present evidence that the stimulation of proliferative and helper T cells by the mutant B6.C.H-2bm1 results from the H-2Kba antigen being recognized in the context of syngeneic Ia determinants. Thus responses to both conventional antigens and allogeneic MHC gene products may proceed via the recognition of antigen in the context of self Ia molecules.     

基于高阶紧致格式的二维不可压NS方程求解

本文介绍了一种基于原始变量的用于求解二维非定常不可压Navier-Stokes方程的高阶紧致格式。这种紧致格式最初是用于计算声学（CAA）的高精度格式,相对于传统的紧致格式,使用该格式的优点在于减少计算量的同时降低了边界模板的处理难度。这种方法建立在非交错网格上,空间离散具有六阶精度。压力Poisson方程基于九基点模板的四阶紧致格式进行离散,超松弛迭代进行求解。时间推进上采用四阶Runge-Kutta方法。为验证该方法的精度和有效性,利用该格式计算了一个具有解析解的问题,以及二维非定常情况下的方腔驱动流动问题,并且和传统的紧致格式进行了计算时间的对比。     

1.5维逆散射级数法消除海上地震勘探资料中的鬼波

针对传统2维逆散射方法计算量大、对观测系统要求高等问题, 将2维逆散射级数方法降维, 得到1.5维逆散射级数消除鬼波的算法。资料处理结果表明, 1.5维方法减少了计算量, 降低了对观测系统的要求, 不需要地下介质信息以及子波估计, 适用于各种复杂的介质情况和低信噪比数据, 可以有效地消除海上地震资料的鬼波, 提高分辨率, 补偿陷波频率处的能量, 提升地震资料的低频成分。     

Systematic method of applying ANN for chemical kinetics reduction in turbulent premixed combustion modeling

A novel method to apply artificial neural network (ANN) for both chemical kinetics reduction and source term evaluation is introduced and tested in direct numerical simulation (DNS) and large eddy simulation (LES) of reactive flows. To gather turbulence affected flame data for ANN training, a new computation-economical method, called 1D pseudo-velocity disturbed flame (PVDF), is developed and used to generate thermo-chemical states independent of the modeled flame. Then a back-propagation ANN is trained using scaled conjugate gradient algorithm to memorize the sample states with reduced orders. The new method is employed in DNS and LES modeling of H₂/air and C₃H₈/air premixed flames experiencing various levels of turbulence. The test result shows that compared to traditional computation with full mechanism and direct integration, this method can obtain quite large speed-ups with adequate prediction accuracy.     

电力系统功角测量方案的研究

电力系统中功角δ和系统的状态有着密切的联系,是判别电力系统稳定性的一个重要参量．例如,发电机的正常和进相运行、系统故障及其恢复过程中,功角的摆动和振荡等,都离不开功角的测量、监视和记录．因此,对功角的测量无论是对电力系统的静态稳定还是暂态稳定都有着极其重要的意义．但多年来,一直无简易的方法实时遥测功角值,而是通过遥测其它各量计算出功角的估计值．笔者提出了两种直接测量功角的方法,并对其测量精度和误差进行了比较分析．因此,利用ＧＰＳ同步时钟可精确地对功角进行测定．     

热力涡轮机械的数字仿真

本文发展了一种积木式的动态仿真程序，它可以仿真由典型设备作任意组合的热力涡轮机械系统。每一个典型设备是一个易于和其它设备相联结的子系统，有一套节省机时的联结方法。本文中所介绍的程序可用于计算非线性系统的大扰动下的过渡过程。文中把催化裂化装置中的能量回收系统作为上述方法的应用实例。     

非均匀网格上求解对流扩散问题的高阶紧致差分方法

基于非均网格上函数的泰勒级数展开,推导出求解一维对流扩散问题的高阶紧致差分格式.对于离散化得到的代数方程组,采用BiCGStab（2）迭代法求解.数值实验表明,该格式对于扩散占优、对流占优及边界层问题都有很好的适应性,对于数值模拟待求物理量的大梯度变化具有很高的分辨率,计算结果明显优于传统的均匀网格上的差分格式.在具体的数值模拟中,可根据实际物理量的变化规律,选取适当的网格生成变换函数,合理地调整非均匀网格的疏密分布,从而获得比在含相同结点数的均匀网络系统中更为精确的数值结果.     

A Deep Learning Method for Chinese Singer Identification

As a subfield of Multimedia Information Retrieval(MIR), Singer IDentification(SID) is still in the research phase. On one hand, SID cannot easily achieve high accuracy because the singing voice is difficult to model and always disturbed by the background instrumental music. On the other hand, the performance of conventional machine learning methods is limited by the scale of the training dataset. This study proposes a new deep learning approach based on Long Short-Term Memory(LSTM) and Mel-Frequency Cepstral Coefficient(MFCC) features to identify the singer of a song in large datasets. The results of this study indicate that LSTM can be used to build a representation of the relationships between different MFCC frames. The experimental results show that the proposed method achieves better accuracy for Chinese SID in the MIR-1 K dataset than the traditional approaches.     

三维激光球杆测量装置及其应用

研制了一种三维激光球杆装置,能方便地测量机床主轴或刀具的空间运行位置,介绍了装置的组成,误差分析和精度较正过程,该装置用于并联机床主轴空间运动位置的测量,实验和测量结果表明具有较高的测量精度和广泛的应用前景。     

滞后的迭代系统的结构与无条件关联稳定性(1)

随着计算机,特别是微型计算机的广泛应用,研究连续动力系统的离散化和工程实际中提出的任意滞后的迭代系统的结构与无条件关联稳定性是很有意义的课题。本文在1986年、1988年刘永清,李在根、叶乐年工作的基础上,采用1957～1960年秦元勋、刘永清创建的方法;给出了在传递滞后与采样周期成整数倍、分数倍与无理数倍时,任意滞后线性定常迭代系统的结构与无条件关联稳定性的充分与必要条件。并把此问题的超越方程求根的问题。化归为代数方程求根问题,还给出了n=1时具体运算过程与结果。对其它非定常、非线性等问题将另文论及。     

Bulk glasses and ultrahard nanoceramics based on alumina and rare-earth oxides

Although often regarded as a network-former in conventional silicate glasses, Al(2)O(3) alone cannot be obtained as a bulk glass. Until now, glasses comprising continuously linked [AlO(x)] polyhedra have been prepared in only a few systems under very fast cooling conditions, which limits their dimensions to a few millimetres. Yet it is desirable to prepare bulk, or monolithic, alumina-rich glasses, with the prospect of superior mechanical, chemical and optical properties. Here we report a novel process for preparing very-high-alumina glasses and nanoscale glass-ceramics. Fully dense bulk articles in net shape are obtained through viscous sintering of glass microbeads. Additional heat treatment of the consolidated glasses leads to fully crystallized transparent glass-converted nanoceramics with a hardness similar to that of alumina. This method avoids the impracticably high applied pressures (more than 1 GPa) that have been required in most cases to prepare nanocrystalline ceramics by sintering, owing to the concurrent nature of densification and grain growth under pressureless conditions. The reported techniques can be extended to form glasses and nanoceramics in other oxide systems that do not include a conventional glass-forming component.     

Applications of cascade multilevel inverters

Cascade multilevel inverters have been developed for electric utility applications. A cascade M-level inverter consists of (M-1)/2 H-bridges in which each bridge‘s dc voltage is supported by its own de ca-pacitor. The new inverter can : ( 1 ) generate almost sinusoidal waveform voltage while only switching one timeper fundamental cycle ; (2) dispense with multi-pulse inverters‘ transformers used in conventional utility in-terfaces and static var compensators; (3) enables direct parallel or series transformer-less connection to medium- and high-voltage power systems. In short, the cascade inverter is much more efficient and suitable for utility applications than traditional multi-pulse and pulse width modulation (PWM) inverters. The authors have experimentally demonstrated the superiority of the new inverter for power supply, (hybrid) electric vehicle (EV) motor drive, reactive power (var) and harmonic compensation. This paper summarizes the features,feasibility, and control schemes of the cascade inverter for utility applications including utility interface of renewable energy, voltage regulation, var compensation, and harmonic filtering in power systems. Analytical,simulated, and experimental results demonstrated the superiority of the new inverters.     

Coupling strength of charge carriers to spin fluctuations in high-temperature superconductors

In conventional superconductors, the most direct evidence of the mechanism responsible for superconductivity comes from tunnelling experiments, which provide a clear picture of the underlying electron-phonon interactions. As the coherence length in conventional superconductors is large, the tunnelling process probes several atomic layers into the bulk of the material; the observed structure in the current-voltage characteristics at the phonon energies gives, through inversion of the Eliashberg equations, the electron-phonon spectral density alpha2F(omega). The situation is different for the high-temperature copper oxide superconductors, where the coherence length (particularly for c-axis tunnelling) can be very short. Because of this, methods such as optical spectroscopy and neutron scattering provide a better route for investigating the underlying mechanism, as they probe bulk properties. Accurate reflection measurements at infrared wavelengths and precise polarized neutron-scattering data are now available for a variety of the copper oxides, and here we show that the conducting carriers (probed by infrared spectroscopy) are strongly coupled to a resonance structure in the spectrum of spin fluctuations (measured by neutron scattering). The coupling strength inferred from those results is sufficient to account for the high transition temperatures of the copper oxides, highlighting a prominent role for spin fluctuations in driving superconductivity in these materials.     

三维非均匀电大目标散射的有限元撕接区域分解法计算

针对非均匀电大目标的散射问题,将有限元撕接区域分解方法(finite element tearing and interconnecting,FETI)应用于非均匀电大目标散射的精确计算模型.该计算模型以各向异性完全匹配吸收层作为吸收边界.数值实验表明,该计算模型比一般吸收边界条件模型精度高.更为重要的是,虽然该模型由于使用各向异性完全匹配吸收层导致计算效率略有降低,但降低十分有限,有限元撕接区域分解方法对于电大尺寸目标的该模型仍然具有很高的计算效率.结果表明,有限元撕接区域分解方法是计算非均匀电大目标散射的一种有效方法.     

Hard-tip, soft-spring lithography

Nanofabrication strategies are becoming increasingly expensive and equipment-intensive, and consequently less accessible to researchers. As an alternative, scanning probe lithography has become a popular means of preparing nanoscale structures, in part owing to its relatively low cost and high resolution, and a registration accuracy that exceeds most existing technologies. However, increasing the throughput of cantilever-based scanning probe systems while maintaining their resolution and registration advantages has from the outset been a significant challenge. Even with impressive recent advances in cantilever array design, such arrays tend to be highly specialized for a given application, expensive, and often difficult to implement. It is therefore difficult to imagine commercially viable production methods based on scanning probe systems that rely on conventional cantilevers. Here we describe a low-cost and scalable cantilever-free tip-based nanopatterning method that uses an array of hard silicon tips mounted onto an elastomeric backing. This method-which we term hard-tip, soft-spring lithography-overcomes the throughput problems of cantilever-based scanning probe systems and the resolution limits imposed by the use of elastomeric stamps and tips: it is capable of delivering materials or energy to a surface to create arbitrary patterns of features with sub-50-nm resolution over centimetre-scale areas. We argue that hard-tip, soft-spring lithography is a versatile nanolithography strategy that should be widely adopted by academic and industrial researchers for rapid prototyping applications.     

Low-energy acoustic plasmons at metal surfaces

Nearly two-dimensional (2D) metallic systems formed in charge inversion layers and artificial layered materials permit the existence of low-energy collective excitations, called 2D plasmons, which are not found in a three-dimensional (3D) metal. These excitations have caused considerable interest because their low energy allows them to participate in many dynamical processes involving electrons and phonons, and because they might mediate the formation of Cooper pairs in high-transition-temperature superconductors. Metals often support electronic states that are confined to the surface, forming a nearly 2D electron-density layer. However, it was argued that these systems could not support low-energy collective excitations because they would be screened out by the underlying bulk electrons. Rather, metallic surfaces should support only conventional surface plasmons-higher-energy modes that depend only on the electron density. Surface plasmons have important applications in microscopy and sub-wavelength optics, but have no relevance to the low-energy dynamics. Here we show that, in contrast to expectations, a low-energy collective excitation mode can be found on bare metal surfaces. The mode has an acoustic (linear) dispersion, different to the dependence of a 2D plasmon, and was observed on Be(0001) using angle-resolved electron energy loss spectroscopy. First-principles calculations show that it is caused by the coexistence of a partially occupied quasi-2D surface-state band with the underlying 3D bulk electron continuum and also that the non-local character of the dielectric function prevents it from being screened out by the 3D states. The acoustic plasmon reported here has a very general character and should be present on many metal surfaces. Furthermore, its acoustic dispersion allows the confinement of light on small surface areas and in a broad frequency range, which is relevant for nano-optics and photonics applications.     

The RCT approach for cell coverage prediction (II): Experiment and early results

We study on the implementation flow of the radio computerized tomography (RCT) prediction method. A case in real cellular mobile radio (CMR) system together with the prediction results are also presented. As shown by the results, the RCT prediction method is marked for its convenience and rapidity, as well as its relative high precision even when the prediction procedure is highly simplified. Since it is developed according to the characteristics of wireless communication environments of our country and has concurrently merits from both statistical and deterministic prediction models, the RCT prediction method is in good agreement with engineering practices in cellular mobile communication in cities at home. Optimized by combining with other techniques, further improvement could be achieved in the stability and precision of the RCT prediction method which now serves as the core part of a software tool for commercial use in CMR system analysis and optimization. Biography: Yu Sheng-bing (1974-), male, Ph. D candidate, research interest: theory analysis and application development of radio propagation in wireless communication systems.     

一种多摄像机视觉系统的标定方法

设计了一种标定多摄像机视觉系统的方法,在多摄像机场景中,放置共线的2点(一维标靶)并使其围绕1个固定端点摆动;通过多摄像机从各自角度同步拍摄,标定出各摄像机含径向畸变系数的内外参数。文中使用的标定物制作简单,解决了标定物各角度的可见性问题,适用于对分布式安放的多台摄像机进行集体标定。     

Fast Multipole BEM for 3-D Elastostatic Problems with Applications for Thin Structures

The fast multipole method (FMM) has been used to reduce the computing operations and memory requirements in large numerical analysis problems. In this paper, the FMM based on Taylor expansions is combined with the boundary element method (BEM) for three-dimensional elastostatic problems to solve thin plate and shell structures. The fast multipole boundary element method (FM-BEM)requires O(N) operations and memory for problems with N unknowns. The numerical results indicate that for the analysis of thin structures, the FM-BEM is much more efficient than the conventional BEM and the accuracy achieved is sufficient for engineering applications.