基于PowerPC和X86的二余度非相似飞控计算机系统设计与实现

为了抑制共模故障和提高飞控计算机系统的可靠性,设计实现了一种二余度非相似的飞控计算机原型系统。首先,给出了该系统的硬件结构和原理,硬件结构包括基于PowerPC和X86处理器构成的主副通道、用于交叉通信的双口RAM、基于FPGA的仲裁模块等部分;其次,系统中的余度控制软件部分主要给出了任务同步和仲裁处理软件模块的软件实现流程图,这是非相似余度的关键问题;最后,在实现实物系统的基础上,开展了一系列的仿真实验,验证了该设计系统的可行性。     

Experiment and analysis of the effect of fine tracking system on the unstable platform in laser communication

Atmospheric turbulence and platform vibration in the space optical communication can cause the offset and jitter of beam, which further result in the fluctuations of received optical power. To resist this effect, a communication system with fine tracking systems in the receiver and transmitter is designed. The system is used in the experiment of laser communication between high-rise buildings over a distance of 3.5 km. After adding a vibration source to the transmitter, the centroids of spots captured by the camera of the transmitter and the optical power of receiver are recorded for the purpose of analysis. When the vibration source works at the designated frequency, a peak appears at the corresponding frequency in the spectrum of the spot centroids and the optical power of receiver. Then the peak disappears once the fine tracking system begins to work. Compared with the condition without the fine tracking system, the minimum value of the optical power of receiver is increased by 5 dB, and the standard deviation is decreased by 30%.     

功率MOSFET零温度系数点的研究

为研究功率MOSFET零温度系数点（Zero Temperature Coefficient,ZTC）,不同栅极电压对温度系数的影响,本文基于阈值电压和迁移率的温度关系并结合功率MOSFET输出特性模型,得到了随栅极电压变化而出现的三种不同温度系数。利用实际测试验证了温度系数的变化规律,结果表明：在小栅极电压条件下,线性区和饱和区均为正温度系数;随栅极电压增大,线性区先进入负温度系数,饱和区仍然为正温度系数,并因此产生了零温度系数点;随栅极电压进一步增大,线性区和饱和区都进入负温度系数。同时根据测试数据,分析了ZTC在不同栅极电压条件下随温度的变化原因,并基于测试数据讨论了迁移率随栅极电压和温度的变化关系。     

A Method for Quantification of the Effects of Size and Geometry on the Microstructure of Miniature Interconnects

Because the heterogeneity of microstructure has significant effects on the material properties of ultrafine interconnects, it should be quantified, to facilitate high-fidelity prediction of reliability. To address this challenge, a method based on autocorrelation and singular value decomposition is proposed for quantitative characterization of microstructure. The method was validated by developing a quantitative relationship between reported microstructure and tensile strength for SnAgCuRE solders reported in the literature. The method was used to study the effects of size and geometry in ultrafine Sn37Pb interconnects on microstructure and von Mises stress, which were obtained simultaneously by coupling a phase-field model with an elastic mechanical model. By use of this method the degree of heterogeneity of the microstructure in relation to preferred growth directions of the phases was quantified by use of a scalar microstructure index. It was found that microstructure heterogeneity increases with decreasing standoff height, and is higher for hourglass-shaped solder joints. The average von Mises stress was found to be positively related to the microstructure index. The strong correlation between microstructure index and average von Mises stress was confirmed by nonlinear regression analysis using an artificial neural network. This indicates that the mechanical behavior of ultrafine interconnects can be predicted more accurately on the basis of the microstructure index.     

A novel low-complexity power allocation algorithm based on the NOMA system in a low-speed environment

For future wireless communication systems, Power Domain Non-Orthogonal Multiple Access (PD-NOMA) using an advanced receiver has been considered as a promising radio access technology candidate. Power allocation plays an important role in the PD-NOMA system because it considerably affects the total throughput and Geometric Mean User Throughput (GMUT) performance. However, most existing studies have not completely accounted for the computational complexity of the power allocation process when the User Terminals (UTs) move in a slow fading channel environment. To resolve such problems, a power allocation method is proposed to considerably reduce the search space of a Full Search Power (FSP) allocation algorithm. The initial power reallocation coefficients will be set to start with former optimal values by the proposed Lemma before searching for optimal power reallocation coefficients based on total throughput performance. Step size and correction granularity will be adjusted within a much narrower power search range while invalid power combinations may be reasonably discarded during the search process. The simulation results show that the proposed power reallocation scheme can greatly reduce computational complexity while the total throughput and GMUT performance loss are not greater than 1.5% compared with the FSP algorithm.     

聚乙二醇阴极修饰层改善聚合物太阳能电池光电性能的研究

通过将聚乙二醇(PEG)掺入活性层制备聚合物太阳 能电池,利用PEG的迁移特性获得阴极修饰层,研 究PEG阴极修饰层对聚合物太阳能电池光电性能的影响。X射线光电子能谱(XPS)分 析表明,掺入活性层中的 PEG迁移到活性层与Al电极之间,形成了阴极缓冲层。吸收光谱、电流密度-电压 特性曲线和外量子 效率谱的分析表明,PEG阴极缓冲层的形成改善了活性层与阴极的界面接触特性, 降低了活性层与电 极之间的能级势垒,有利于载流子传输,因此显著地改善了聚合物太阳能电池的光电性能, 使得器件的开 路电压Voc、短路电流密度Jsc和填充因子(FF)都有明显提高。当P3HT:PCBM 活性层中掺入体积比为0.5%的PEG时,聚合物太阳能电池的能量转换 效率(P CE)最高,达到了3.07%,比未掺杂PEG的参考器件提 高了38.5%。     

Evanescent-Coupled Ge p-i-n Photodetectors on Si-Waveguide With SEG-Ge and Comparative Study of Lateral and Vertical p-i-n Configurations

Si-waveguide-integrated lateral Ge p-i-n photodetectors using novel Si/SiGe buffer and two-step Ge-process are demonstrated for the first time. Comparative analysis between lateral Ge p-i-n and vertical p-Si/i-Ge/n-Ge p-i-n is made. Light is evanescently coupled from Si waveguide to the overlaying Ge- detector, achieving high responsivity of 1.16 A/W at 1550 nm with f_{3 dB} bandwidth of 3.4 GHz for lateral Ge p-i-n detector at 5 V reverse bias. In contrast, vertical p-Si/i-Ge/n-Ge p-i-n has lower responsivity of 0.29 A/W but higher bandwidth of 5.5 GHz at -5 V bias. The higher responsivity of lateral p-i-n detectors is attributed to smaller optical mode overlap with highly doped Ge region as in vertical p-i-n configuration.     

“网河”非稳定流模型在洪泛区洪流演进中的应用

将宽阔的溃口洪水行进范围概化成由河道水塘和联系河道组成的网河,用一维非恒定流方程组来描述主要的水充运动状态,通过简化和变换,改写成统一的四点线性隐式有限差的方程组形式,再用双消除法求解,全部计算在微机上实现,参数用实测资料率定,成果具有较好的精度,在一定程度上反映出横水流方向的水位流量变化,为防洪水利计算开辟一条新的途径。     

二维信号Wigner分布的数值计算

Wigner分布函数作为一个同时表达空间 频率变量的函数 ,在光学领域中有着广泛的应用。提出了二维信号Wigner分布函数的数值计算方法 ,并模拟了一些二维实信号 (圆孔 ,圆环和方孔 )的Wigner分布函数     

一种改进的SLNR多用户MIMO预编码方法

MU-MIMO系统中,基站在相同的时间与频率资源上与多个用户进行通信,导致不同用户之间存在同信道干扰。SLNR预编码方法将联合优化的问题转化成若干个独立优化问题,在系统性能与复杂度之间取到很好的折中。提出了一种改进的SLNR预编码方法,该方法在泄漏信号的基础上,考虑干扰信号的影响,对预编码向量进行迭代优化。仿真结果表明,与SLNR方法相比,该方法在理想CSI和非理想CSI情况下都可以进一步提高系统的吞吐量。