THE DESIGNS OF HIGH EFFICIENCY LAUNCHER OF QUASI-OPTICAL MODE CONVERTER FOR HIGH POWER GYROTRONS

A high efficiency launcher of quasi-optical (QO) mode converters for high power gyrotrons have been designed and tested. A helical cut launcher radiates the RF power via its straight cut onto the first phase correcting mirror. The launchers have been optimized for the TE_31.8 mode at 170 GHz and TE_22.6 mode at 110 GHz by numerically optimizing a launcher surface. The helical cut of the launcher has been optimized by taking the taper angle into account. Further more, the amplitude of the surface perturbation have been optimized for improved focusing in order to reduce the diffraction losses at the helical cut. Low power measurement shows a good agreement with the design. High efficiency characteristics of the design have also been calculated on the assumption of frequency downshift due to the thermal expansion of the cavity and stepwise frequency tuning by changing the operating mode. Besides, the possibility of high efficiency launcher for higher mode is discussed, and these results give the prospect to high efficiency long pulse gyrotrons.     

一维细胞神经网络的完全稳定性分析

运用Lyapunov函数方法,讨论了一维细胞神经网络模型的完全稳定性问题,给出了四组使模型具有完全稳定的充分条件.     

现代优化计算方法在蛋白质结构预测中的应用

现代优化计算方法在蛋白质结构预测中占有重要地位.简要地介绍了模拟退火算法,遗传算法,人工神经网络和图论算法在蛋白质结构预测中的应用.对国内外近年来应用这些算法,特别是在蛋白质构象搜索问题中,解决蛋白质结构预测的研究作了回顾,并分析、比较了这几种算法的效果和特点.     

Non-Parametric Regression Methods

The nature of the financial time series is complex, continuous interchange of stochastic and deterministic regimes. Therefore, it is difficult to forecast with parametric techniques. Instead of parametric models, we propose three techniques and compare with each other. Neural networks and support vector regression (SVR) are two universally approximators. They are data-driven non parametric models. ARCH/GARCH models are also investigated. Our assumption is that the future value of Istanbul Stock Exchange 100 index daily return depends on the financial indicators although there is no known parametric model to explain this relationship. This relationship comes from the technical analysis. Comparison shows that the multi layer perceptron networks overperform the SVR and time series model (GARCH).     

等效可变孔径远场法测量单模光纤模场半径

本文提出等效远场的可变孔径方法,由测量光纤远场累积分布求得模场半径。并且利用Petermann关于模场半径的新定义,推出了一种对非高斯模场分布也适用的求模场半径的方法。实验表明,这种测量方法简便可靠,在不同的制备端面下,重复测量的标准误差小于0.04μm。     

Quadratic distances for capacity and bi-capacity approximation and identification

The application of multi-attribute utility theory based on the Choquet integral requires the prior identification of a capacity if the utility scale is unipolar, or of a bi-capacity if the utility scale is bipolar. In order to implement a minimum distance principle for capacity or bi-capacity approximation or identification, quadratic distances between capacities and bi-capacities are studied. The proposed approach, consisting in solving a strictly convex quadratic program, has been implemented within the GNU R kappalab package for capacity and nonadditive integral manipulation. Its application is illustrated on two examples.      

基于Internet/Intranet的大学物理实验系统

利用计算机技术、传感技术和校园网，将现代教育技术引入大学普通物理实验教学中，实现实验数据的自动采集，基于网络的实验预习和实验报告写作、递交、批改及管理等功能．     

Lorentz-Covariant Loop Quantum Gravity

We review the Lorentz-covariant approach to loop quantum gravity. This approach solves the Immirzi parameter problem occurring in the standard loop approach based on the SU(2) gauge group. We show that there exists a unique loop quantization preserving all the classical symmetries at the quantum level and that the results obtained with it, such as the area operator spectrum, are independent of the Immirzi parameter. The standard SU(2) approach violates the diffeomorphism invariance and is therefore an incorrect quantization of gravity.     

Widrow-cellular neural network and optoelectronic implementation

A new type of optoelectronic cellular neural network has been developed by providing the capability of coefficients adjusment of cellular neural network (CNN) using Widrow based perceptron learning algorithm. The new supervised cellular neural network is called Widrow-CNN. Despite the unsupervised CNN, the proposed learning algorithm allows to use the Widrow-CNN for various image processing applications easily. Also, the capability of CNN for image processing and feature extraction has been improved using basic joint transform correlation architecture. This hardware application presents high speed processing capability compared to digital applications. The optoelectronic Widrow-CNN has been tested for classic CNN feature extraction problems. It yields the best results even in case of hard feature extraction problems such as diagonal line detection and vertical line determination.     

正交变换在重积分中某些应用

正交变换是代数学的基本内容 ,其用途十分广泛 .重积分的计算往往存在技术性的困难 ,若利用“正交变换”的有关理论去解决某些重积分的计算问题是颇有功效的 .本文将以“正交变换”为工具 ,简洁的处理重积分的某些问题