基于结点基函数的时域有限元方法仿真微波无源器件

在仿真微波无源器件时,有限元法(FEM)对多个频率点需重复计算,且很难精确模拟非线性器件及非线性介质,时域有限差分法(FDTD)需采用规则的网格剖分,不能很好地模拟具有弯曲表面的模型,而时域有限元法则可以解决这些难题,因此TDFEM在微波无源器件仿真中的应用具有重要的研究意义。采用结点基函数可以简化TDFEM的复杂度,并大幅度缩小系数矩阵维数,从而缩短计算所需时间和内存,因此,本文研究了基于结点基函数的TDFEM仿真微波无源器件。数值算例显示了基于结点基函数的TDFEM能够简单、正确、有效地仿真微波无源器件。

基于混沌序列的最大和分散式协调算法

大规模信息系统的协调问题正成为新兴的研究热点,最大和协调算法显示了在该领域的应用前景,然而其收敛速度和鲁棒性有待进一步提高。针对以上问题,提出一种可用于合作系统中的基于混沌的分散式信息传递算法,并通过图形着色问题验证了该算法的有效性。在该算法中,利用混沌序列丰富的时空动态性,产生最大和协调算法的结点信息。然后依据更新规则对结点信息进行交互,完成算法的迭代。通过与传统最大和协调算法和遗传算法的对比实验结果分析,该算法在收敛速度及鲁棒性方面有着更好的表现。

二维TM波时域非连续伽略金算法理论数值通量研究

采用数值通量的方式进行场量交互是时域非连续伽略金(Discontinuous Galerkin Time Domain,DGTD)算法区别于时域有限元(Finite Element Time Domain,FETD)方法的主要方面.从二维TM情形弱解方程出发,讨论了当前三角形单元和相邻单元进行场量交互时数值通量物理意义和不同形式.结合数值通量和弱解方程得到了DGTD算法的迭代计算式.给出了线元辐射和双线元干涉的数值算例,算例结果表明了文中方法的正确性.

Influences of a Side-Coupled Triple Quantum Dot on Kondo Transport Through a Quantum Dot

Kondo transport properties through a Kondo-type quantum dot （QD） with a side-coupled triple-QD structure are systematically investigated by using the non-equilibrium Green＇s function method. We firstly derive the formulae of the current, the linear conductance, the transmission coefficient, and the local density of states. Then we carry out the analytical and numerical studies and some universal conductance properties are obtained. It is shown that the number of the conductance valleys is intrinsically determined by the side-coupled QDs and at most equal to the number of the QDs included in the side-coupled structure in the asymmetric limit. In the process of forming the conductance valleys, the side-coupled QD system plays the dominant role while the couplings between the Kondo-type QD and the side-coupled structure play the subsidiary and indispensable roles. To testify the validity of the universal conductance properties, another different kinds of side-coupled triple-QD structures are considered. It should be emphasized that these universal properties axe applicable in understanding this kind of systems with arbitrary many-QD side structures.

Pole-Based Approximation of the Fermi-Dirac Function

Two approaches for the efficient rational approximation of the Fermi-Dirac function are discussed： one uses the contour integral representation and conformal real〉 ping, and the other is based on a version of the multipole representation of the Fermi-Dirac function that uses only simple poles. Both representations have logarithmic computational complexity. They are of great interest for electronic structure calculations.

A partition-of-unity based three-node triangular element with continuous nodal stress using radial-polynomial basis functions

A partition-of-unity （PU） based ＂FE-Meshfree＂ three-node triangular element （Trig3-RPIM） was recently developed for linear elastic problems. This Trig3-RPIM element employs hybrid shape functions that combine the shape functions of three-node triangular element （Trig3） and radial-polynomial basis functions for the purpose of synergizing the merits of both finite element method and meshfree method. Although Trig3-RPIM element is capable of obtaining higher accuracy and convergence rate than the Trig3 element and four-node iso-parametric quadrilateral element without adding extra nodes or degrees of freedom （DOFs）, the nodal stress field through Trig3-RP1M element is not continuous and extra stress smooth operations are still needed in the post processing stage. To further improve the property of Trig3-RPIM element, a new PU-based triangular element with continuous nodal stress, called Trig3-RPIMcns, is developed. Numerical examples including several linear, free vibration and forced vibration test problems, have confirmed the correctness and feasibility of the proposed Trig3-RPIMcns element.