The Newton Bracketing Method for Convex Minimization

An iterative method for the minimization of convex functions f :ⁿ , called a Newton Bracketing (NB) method, is presented. The NB method proceeds by using Newton iterations to improve upper and lower bounds on the minimum value. The NB method is valid for n = 1, and in some cases for n > 1 (sufficient conditions given here). The NB method is applied to large scale Fermat–Weber location problems.     

A half-moment model for radiative transfer in a 3D gray medium and its reduction to a moment model for hot, opaque sources

We present in this paper a new 3D half-moment model for radiative transfer in a gray medium, called the model, which uses maximum entropy closure. This model is a generalization to 3D of the 1D version recently proposed in (J. Comp. Phys. 180 (2002) 584). The direction space Ω is divided into two pieces, Ω⁺ and Ω^-, in a dynamical way by the plane perpendicular to the total radiative flux, and the half moments are defined from these subspaces. The model closure and the integrations of the radiative transfer equation performed on the moving Ω^± spaces are detailed. 1D planar results, which have motivated the extension of the model of (J. Comp. Phys. 180 (2002) 584) to multi-dimensions, are shown. These results are very good. The model is thereafter derived for 3D spherically symmetric geometry, where the correctness of the non-trivial border terms can be checked. Two 3D spherically symmetric problems are numerically solved in order to show the accuracy of the closure and the role of the border terms. Once again, compared to the solution obtained with a ray tracing solver, results are very good. From the 3D half-moment model, a new moment model, called , is derived for the particular case of a 3D hot and opaque source radiating into a cold medium, for applications such as simulations of stellar atmospheres and fires. Two-dimensional numerical results are presented and compared to those obtained solving the RTE and with other moment models. They demonstrate the very good accuracy of the model, its good convergence properties, and better prediction compared to all other existing moment models in its domain of applicability.     

利用偏振控制器连续调节光纤环镜的反射率

研究了一种新型光纤环镜(FLM)的原理与特性，这种FLM由在普通光纤环镜中插入光纤型偏振控制器(PC)构成.通过等效光路分析建立了该FLM的理论模型，并对其反射特性进行了数值模拟.研究表明，通过改变PC的状态，即改变其双折射效应的快轴取向或强度，可连续调节FLM的反射率，反射率谱具有宽带特性，主要受光纤耦合器工作带宽的影响.此外，对FLM的反射特性还进行了实验研究.实验结果也证实，通过调节PC状态，FLM反射率可在其最大和最小值之间连续调节，实验测得FLM最大和最小反射率分别可达93%和2%.根据PC双折 关键词： 光纤环镜 偏振控制器 光纤耦合器 反射率     

Analysis of Coupling and Radiation Properties for an Improved Dielectric Waveguide Directional Coupler Utilizing a Tapering Left-Handed Slab Core

The coupling and radiation characteristics of an improved dielectric waveguide directional coupler with a tapering left-handed slab core located in the coupling region are analyzed by the staircase approximation method, which combines the building block approach and multi-mode network theory with a rigorous mode-matching procedure. Particular attentions are directed toward the radiation characteristics under different structural parameters of the coupler. Numerical results indicate that the proposed coupler is very promising by shortening the coupling length significantly while keeping the radiation loss at a low level.     

A numerical model of multimode fiber PON frequency response

Frequency response of passive optical network (PON) based on multimode fibers is investigated. The network comprises fibers, connectors and splitters/couplers. It is shown that due to mode filtering at splitters, the frequency response is different for different network nodes in otherwise symmetrical network.     

一种双频微波功率合成器的设计与实现

一种新型的双频微波功率合成器，通过阶跃阻抗低通滤波器和平行耦合线带通滤波器构成的微带电路实现。该微波功率合成器实现了915MHz和2．5GHz两个非相干频率微波信号的功率合成。同时通过附加工作在915MHz和2．5GHz的20dB微带定向耦合器，就可以构建实验系统进行微波功率合成的测量。实验测量结果与数值模拟结果吻合，双频微波功率合成器和定向耦合器的性能达到了要求，可用于注入式微波器件效应的研究。     

RELIABILITY ASSESSMENT OF STRUCTURAL SYSTEMS BASED ON DIRECTIONAL VECTOR SIMULATION TECHNIQUE

The new improved directional vector simulation method for analyzing the reliability of structural systems failure probability is researched. This paper also points out the defects of general directional vector simulation, and gives rise to a new higher accuracy approximate integral formula of structural systems failure probability. A new geometric meaning of characteristic function is obtained. A new simple method of generating uniformly distributed random vector samples inn-dimensional unit hyper-spherical surface is put forward and strictly proved. This method is easy to put into practice. Numerical examples are given to show the applicability and effectiveness of the suggested approach to structural systems reliability problems. Supported by Chinese Postdoctor Fund([1998]6, 23).     

Crosstalks of two-waveguide and three-waveguide directional couplers

Crosstalks of two-waveguide and three-waveguide directional couplers are analyzed mathematically based on the relationship between the coupled mode and the normal mode. Numerical techniques such as the finite-difference method and the beam propagation method are employed to confirm the validity of the derived expressions. The calculation results show that two-waveguide directional coupler is superior to any type of three-waveguide couplers from the practical viewpoint of the crosstalk and the coupling length.     

Fluids with highly directional attractive forces. IV. Equilibrium polymerization

We investigate approximation methods for systems of molecules interacting by core repulsion and highly directional attraction due to several attraction sites. The force model chosen imitates a chemical bond by providing for bond saturation when binding occurs. The dense fluid is an equilibrium mixture ofs-mers with mutual repulsion. We use a previously derived reformulation of statistical thermodynamics, in which the particle species are monomeric units with a specified set of attraction sites bonded. Thermodynamic perturbation theory (TPT) and integral equations of two types are derived. The use of TPT is illustrated by explicit calculation for a molecular model with two attraction sites, capable of forming chain and ring polymers. Successes and defects of TPT are discussed. The integral equations for pair correlations between particles of specified bonding include calculation of self-consistent densities of species. Methods of calculating thermodynamic properties from the solutions of integral equations are given.Supported by the NSF under grant No. CHE-82-11236.     

Characteristics of a Grating-Folded Optical Directional Coupler

We developed a new optical add/drop multiplexer composed only of a grating-folded directional coupler. We found that in this multiplexer, the power drop becomes maximal when the grating-coupling strength is equal to that of directional coupling. At this maximum power, both power transfer between the waveguides and power reflection take place simultaneously at the grating. The power drop approaches 1 when the grating length increases.