语音合成ASIC的原理与设计方法

本文阐述了语音合成专用集成电路(ASIC)的数字信号处理原理.采用波形自适应增量调制编码方法,研究并解决了语音信号的压缩与加密.用半定制标准单元CAD设计方法,CMOS工艺,完成了32K ROM 6秒语音ASIC芯片的设计.     

The Drazin inverses of products and differences of orthogonal projections

In this note, the Drazin inverses of products and differences of orthogonal projections on a Hilbert space are established.     

四阶方程两点边值问题Hermite有限元解的渐近展式与外推

1引言有限元解的渐近展式是提高微分方程数值解精度的重要工具,比如亏量校正和外推就是建立在有限元解的渐近展式的基础之上．许多作者对此进行了大量的研究(见[1]-[4]),特别是文[1],提出了在研究有限元解的渐近展式中十分有用的能量嵌入技巧．本文利用能量嵌入定理得到了四阶方程两点边值问题Hermite有限元解及其二阶平均导数的渐近展式,进一步我们还讨论了它们的Richardson外推公式．考虑四阶方程两点边值问题     

LiNbO3∶Cr∶Cu晶体吸收特性及非挥发全息存储研究

研究了LiNbO3∶Cr∶Cu晶体的吸收特性,发现LiNbO3∶Cr∶Cu（含0.14 wt.% Cr2O3 和 0.011 wt.% CuO）晶体存在两个明显的吸收峰,中心波长分别位于480 nm和660 nm; 随着Cr的含量逐渐减小,Cu的含量逐渐增大,短波段不存在明显吸收峰,掺Cr的含量越大,中心波长在660 nm处的吸收越大；633 nm红光虽然位于中心波长为660 nm的吸收峰内,但它无助于光折变过程.分别采用390 nm紫外光和488 nm蓝光作为敏化光,514 nm绿光作为记录光的记录方案,实现了非挥发全息记录,掺入适量的Cr( 比如NCr=2.795×1025 m－3,NCr/ NCu=1)有助于全息记录性能的提高.     

准二维无序系统的电子结构

对形如N_t×N_l型准二维无序系统，只考虑格点之间的最近邻跳跃积分，采用特殊的格点编号方案，在单电子近似下，系统的哈密顿量可表示为简明对称矩阵，借助豪斯荷尔德变换将其约化为对称三对角矩阵，再利用负本征值理论及传输矩阵等方法,对系统态密度、局域长度及电导等电子结构特性进行数值计算. 重点研究了准一维四平行链和五平行链无序系统， 将结果与一维单链、准一维双链及三链系统进行对比，发现随维度的增加，系统的能带有所展宽，能态密度分布发生很大的变化，其峰值数量呈偶数规律增加. 并且在能带中心处存在有局域长度大于系统大小的扩展态，处于这些态下的系统具有较大电导. 从单链到多链，相当于扩大了系统的关联范围，使系统出现了类似非对角长程关联的行为. 关键词： 准二维无序系统 态密度 局域长度 电导     

Large-scale magnetic structures of coronal mass ejections

Based on the studies on the source regions of a group of coronal mass ejections, we have identified two types of large-scale magnetic structures, and suggested that they are intrinsic components of solar magnetism, their destabilization, expansion, and eruption into the interplanetary space are the basic physical processes which lead to the coronal mass ejections. These two types of large-scale structures are giant magnetic loops connecting the two active belts on the opposite hemispheres of the Sun, and the giant filaments (filament channels) and their related magnetic structures. The latter often appear as two parallel rows of sunspots and plage fields, which align side by side in the full disk daily and synoptic magnetograms. The magnetic neutral lines of these large-scale structures are usually longer than 50 heliographic degrees. We name this type of structure “super A configuration”. Sometimes, they are shown as very long filaments and related large-scale magnetic fields. As these magnetic structures are of very large scale, they extend to a great altitude into the corona, they are not easily recognized in magnetic field observations which are usually aimed at solar flare studies. To identify these large-scale structures becomes a key to understanding and predicting coronal mass ejections.     

Self-consistent calculation of the lasing eigenmode of thedielectrically apertured Fabry-Perot microcavity with idealized ordistributed Bragg reflectors

Optical confinement in the dielectrically apertured Fabry-Perot microcavity is investigated theoretically. The apertured region is first treated as embedded in an idealized planar waveguide to show that the confined eigenmode's resonant frequency can cut off parasitic waveguide modes existing outside the aperture, and lead to three-dimensional optical confinement. For more realistic cavities with nonunity mirror reflectivities, self-consistent calculations of the eigenmode characteristics are performed for the limit of an optically thin aperture to derive the lowest order confined eigenmode frequency, threshold susceptibility, and mode profile. The analysis is then extended to treat dielectric cavities based on Bragg reflectors     

Integral representations of harmonic functions in half spaces

In this paper, using a modified Poisson kernel in an upper half-space, we prove that a harmonic function u(z) in a upper half space with its positive part u⁺(x)=max{u(x),0} satisfying a slowly growing condition can be represented by its integral in the boundary of the upper half space, the integral representation is unique up to the addition of a harmonic polynomial, vanishing in the boundary of the upper half space and that its negative part u⁻(x)=max{−u(x),0} can be dominated by a similar slowly growing condition, this improves some classical result about harmonic functions in the upper half space.