Robust design of an optical router based on a tapered side-coupled integrated spaced sequence of optical resonators

A novel (to our knowledge) scheme of an optical router/switch element, composed of a tapered side-coupled integrated spaced sequence of optical resonators, is proposed. It is based on a modified design of the ring sequence in which the resonance conditions are set by the single ring resonance and by the coherent feedback of the sequence of rings. This double condition yields robustness against fabrication defects, dense routing capability, and high switching efficiency.     

参数不确定时延超混沌系统的脉冲同步方法研究

针对一类参数不确定的异构时延超混沌系统,采用响应系统与驱动系统状态变量误差的线性反馈作为脉冲控制信号,基于Lyapunov稳定性理论实现了超混沌系统的脉冲同步,给出了渐近稳定条件,并构造了一个具有时延的数字保密通信系统.该方案保密性高,鲁棒性强.数值仿真表明系统能快速达到同步,应用混沌密码序列对数字信号进行加密后,信息信号可以在接收端有效地恢复出来. 关键词： 时延 脉冲同步 数字保密通信     

基于改进的隐马尔可夫和神经网络混合模型的语音识别

研究了一种非齐次隐马尔可夫模型（Inhomogeneous Hidden Markov Model），然后将自组织特征映射神经网络与这种非齐次隐马尔可夫模型相结合，训练出抗噪声的HMM模型，并应用该混合模型进行语音识别。实验结果表明，该模型适合于对噪声背景下的语音进行识别。该模型具有更好的抗噪鲁棒性，在信噪比较低的情况下（5dB-10dB），识别率可以提高5％左右。     

求解流动和传热问题的一种新的全隐算法-CLEAR(下)

本文上篇阐述了CLEAR算法的推导过程和计算步骤,本文下篇通过五个二维不可压缩流动和传热的算例,对CLEAR算法和SIMPLER算法进行了比较,比较的内容为,在相同的收敛条件下, CLEAR算法和SIMPLER算法收敛所需的迭代次数的比值和对应的CPU时间的比值,以及这两个比值和时步倍率的关系,从而进一步研究了CLEAR算法的健壮性。计算结果表明,CLEAR算法可以在很大程度上加速迭代收敛,就所比较的算例而言,其可以节省迭代次数31％-85％,节省CPU时间17％-78％,而且该算法的健壮性可以通过引入第二松弛因子而得以提高。     

Lattice Boltzmann simulation of flow around two,three and four circular cylinders in close proximity

Cross-flows around two, three and four circular cylinders in tandem, side-by-side, isosceles triangle and square arrangements are simulated using the incompressible lattice Boltzmann method with a second-order accurate curved boundary condition at Reynolds number 200 and the cylinder center-to-center transverse or/and longitudinal spacing 1.5D, where D is the identical circular cylinder diameter. The wake patterns, pressure and force distributions on the cylinders and mechanism of flow dynamics are investigated and compared among the four cases. The results also show that flows around the three or four cylinders significantly differ from those of the two cylinders in the tandem and side-by-side arrangements although there are some common features among the four cases due to their similarity of structures, which are interesting, complex and useful for practical applications. This study provides a useful database to validate the simplicity, accuracy and robustness of the Lattice Boltzmann method.

     

带有层级结构的复杂网络级联失效模型

针对现实世界的网络中普遍存在的层级结构建立一个级联失效模型, 该模型可用于优化金融、物流网络设计. 选择的层级网络模型具有树形骨架和异质的隐含连接, 并且骨架中每层节点拥有的分枝数服从正态分布. 级联失效模型中对底层节点的打击在不完全信息条件下进行, 也即假设打击者无法观察到隐含连接. 失效节点的负载重分配考虑了层级异质性, 它可以选择倾向于向同级或高层级完好节点分配额外负载. 仿真实验表明, 层级网络的拓扑结构随连接参数变化逐渐从小世界网络过渡到随机网络. 网络级联失效规模随隐含连接比例呈现出先增加后降低的规律. 负载重分配越倾向于高层级节点, 网络的抗毁损性越高. 同时, 由于连接参数会改变隐含连接在不同层级之间的分布, 进而对网络的抗毁损性产生显著影响, 为了提高网络抗毁损能力, 设计网络、制定管理控制策略时应合理设定连接参数. 关键词： 复杂网络 级联失效 层级结构     

扰动声线声压敏感核用于浅水前向散射小目标定位

针对收发合置声呐在浅水信道中对小目标的定位能力受强混响制约的问题,开展了利用目标前向散射的声屏障实验研究,通过计算扰动声线的声压敏感核实现了在自然环境中、非等声速条件下的小目标定位。扰动声线的声压敏感核反映的是因目标进入探测区域而扰动到的声线所包含的目标位置信息.首先,从众多本征声线中确定可以用于目标定位的扰动声线;然后,利用扰动声线的声压敏感核的空间特性实现定位。将该定位算法用于湖试数据,实现了利用20~28 kHz的线性调频信号对直径0.4 m目标球的定位。并且通过比较不同参数假设下的计算结果,发现该定位算法对于目标材质和尺寸的失配均具有较好的鲁棒性.      

Control performance and robustness of an active noise control system for uncertain primary sound fields

In this work, the control performance and robustness of an active noise control system subjected to uncertain primary sound fields are investigated. For this purpose, the performance index, residual potential energy in a desired quiet zone, is derived as a function of sound field variables, quiet zone variables, and control system variables. In the presence of uncertainty, typical measures of the robustness and performance of a control system, maximum, minimum, mean, and variance of the performance index are derived theoretically. In addition, based on the least-squares orthogonality principle, the condition for implementing the best-oriented control system, which is robust and can maximize the control performance by using a given number of control sources and sensors, is investigated. Finally, numerical simulations are carried out to illustrate and verify the proposed theory.     

Condition number as a measure of noise performance of diffusion tensor data acquisition schemes with MRI

Diffusion tensor mapping with MRI can noninvasively track neural connectivity and has great potential for neural scientific research and clinical applications. For each diffusion tensor imaging (DTI) data acquisition scheme, the diffusion tensor is related to the measured apparent diffusion coefficients (ADC) by a transformation matrix. With theoretical analysis we demonstrate that the noise performance of a DTI scheme is dependent on the condition number of the transformation matrix. To test the theoretical framework, we compared the noise performances of different DTI schemes using Monte-Carlo computer simulations and experimental DTI measurements. Both the simulation and the experimental results confirmed that the noise performances of different DTI schemes are significantly correlated with the condition number of the associated transformation matrices. We therefore applied numerical algorithms to optimize a DTI scheme by minimizing the condition number, hence improving the robustness to experimental noise. In the determination of anisotropic diffusion tensors with different orientations, MRI data acquisitions using a single optimum b value based on the mean diffusivity can produce ADC maps with regional differences in noise level. This will give rise to rotational variances of eigenvalues and anisotropy when diffusion tensor mapping is performed using a DTI scheme with a limited number of diffusion-weighting gradient directions. To reduce this type of artifact, a DTI scheme with not only a small condition number but also a large number of evenly distributed diffusion-weighting gradients in 3D is preferable.     

A consistent and stabilized continuous/discontinuous Galerkin method for fourth-order incompressible flow problems

This paper presents a new consistent and stabilized finite-element formulation for fourth-order incompressible flow problems. The formulation is based on the C⁰-interior penalty method, the Galerkin least-square (GLS) scheme, which assures that the formulation is weakly coercive for spaces that fail to satisfy the inf-sup condition, and considers discontinuous pressure interpolations. A stability analysis through a lemma establishes that the proposed formulation satisfies the inf-sup condition, thus confirming the robustness of the method. This lemma indicates that, at the element level, there exists an optimal or quasi-optimal GLS stability parameter that depends on the polynomial degree used to interpolate the velocity and pressure fields, the geometry of the finite element, and the fluid viscosity term. Numerical experiments are carried out to illustrate the ability of the formulation to deal with arbitrary interpolations for velocity and pressure, and to stabilize large pressure gradients.     

"量子相干性跟踪控制"的鲁棒性分析

文献[Quantum Information and Computation,2005,5(4):350-363]提出量子相干性跟踪控制策略以保持量子位的相干性,该策略本质是开环控制,而开环控制的缺陷就是严重依赖于精确地了解系统的初始条件和模型参数。然而,我们对于系统的知识总是存在不确定性的。本文从鲁棒性的角度分析了量子相干性跟踪控制策略的局限性。首先,我们对该方法的鲁棒性进行了理论分析;接着用仿真实例定量说明了:量子初始条件和模型参数的精度对保持相干性的影响。根据相干性的精度要求,我们可以利用仿真分析的手段,来确定对模型参数和初始状态必须满足的精度条件。总之,我们在应用中必须谨慎采用量子相干性跟踪控制策略。     

Integrative fitting of absorption line profiles with high accuracy,robustness, and speed

The principle of the integrative evaluation of absorption line profiles relies on the numeric integration of absorption line signals to retrieve absorber concentrations, e.g., of trace gases. Thus, it is a fast and robust technique. However, previous implementations of the integrative evaluation principle showed shortcomings in terms of accuracy and the lack of a fit quality indicator. This has motivated the development of an advanced integrative (AI) fitting algorithm. The AI fitting algorithm retains the advantages of previous integrative implementations—robustness and speed—and is able to achieve high accuracy by introduction of a novel iterative fitting process. A comparison of the AI fitting algorithm with the widely used Levenberg–Marquardt (LM) fitting algorithm indicates that the AI algorithm has advantages in terms of robustness due to its independence from appropriately chosen start values for the initialization of the fitting process. In addition, the AI fitting algorithm shows speed advantages typically resulting in a factor of three to four shorter computational times on a standard personal computer. The LM algorithm on the other hand retains advantages in terms of a much higher flexibility, as the AI fitting algorithm is restricted to the evaluation of single absorption lines with precomputed line width. Comparing both fitting algorithms for the specific application of in situ laser hygrometry at 1,370 nm using direct tunable diode laser absorption spectroscopy (TDLAS) suggests that the accuracy of the AI algorithm is equivalent to that of the LM algorithm. For example, a signal-to-noise ratio of 80 and better typically yields a deviation of <1 % between both fitting algorithms. The properties of the AI fitting algorithm make it an interesting alternative if robustness and speed are crucial in an application and if the restriction to a single absorption line is possible. These conditions are fulfilled for the 1,370 nm TDLAS hygrometry at the aerosol and cloud chamber aerosol interactions and dynamics in the atmosphere (AIDA)—a unique large-scale facility to study atmospheric processes. The robustness of the AI fitting algorithm has been validated for typical AIDA conditions encompassing strong transmission fluctuations during the formation of droplet or ice clouds inside AIDA. Under these conditions, the stability of the AI algorithm remained virtually unaffected. Thus, the AI algorithm presents an alternative technique for a fast, reliable, and accurate online data evaluation of the humidity measurements at AIDA.     

On the validity and improvement of the ultrasonic pulse-echo immersion technique to measure real attenuation

A fundamental assumption embraced in conventional use of the ultrasonic pulse-echo immersion technique to measure attenuation in solid materials is revisited. The cited assumption relies on perfect and immutable adhesion at the water to sample interface, a necessary condition that allows calculating the reflection coefficient at any interface from elastic wave propagation theory. This parameter is then used to correct the measured signal and obtain the real attenuation coefficient of the sample under scrutiny. In this paper, cases in which the perfectly cohesive interfacial condition is not satisfied are presented. It is shown also that in those cases, the repeatability of the conditions at the interface is always uncertain. This implies that the reflection coefficients are unknown, even when density is known. A new method of simultaneously measuring the reflection coefficients for both exposed interfaces that are normal to the transducer, and the attenuation coefficient of the specimen is developed and is presented here. The robustness of the new method is proven, as we demonstrate that the proper value of attenuation is achieved independently of the continuously varying interfacial conditions of these non-ideal cases.     

A High-Order NVD/TVD-Based Polynomial Upwind Scheme for the Modified Burgers' Equations

A bounded high order upwind scheme is presented for the modified Burgers' equation by using the normalized-variable formulation in the finite volume framework. The characteristic line of the present scheme in the normalized-variable diagram is designed on the Hermite polynomial interpolation. In order to suppress unphysical oscillations, the present scheme respects both the TVD (total variational diminishing) constraint and CBC (convection boundedness criterion) condition. Numerical results demonstrate the present scheme possesses good robustness and high resolution for the modified Burgers' equation.     

负荷作用下相依网络中的级联故障

研究负荷作用下相依网络中的级联故障具有重要的现实意义, 可为提高相依网络的鲁棒性提供参考. 构建了双层相依网络级联故障模型, 主要研究了外部度和内部度对负荷贡献比、耦合因素、层内度-度相关性对相依网络级联故障的影响. 研究表明, 当外部度和内部度对负荷贡献比达到一定值时, 相依网络抵抗级联故障的鲁棒性最强. 而耦合因素的影响是多方面的, 为了达到较高鲁棒性, 建议采用异配耦合方式和尽可能大的平均外部度, 并尽量使外部度保持均匀分布. 另外, 与不考虑负荷作用时相反, 当表征层内度-度相关性的相关系数越大时, 其抵抗级联故障的能力越强.     

一种新的简化气动BGK格式

介绍了气动BGK格式的最新发展,在此基础上对原格式进行了适当简化.数值验证表明简化后的格式不但保留了原格式的强健性和自动满足熵条件等优点,而且算法更加简明,计算量更小.     

多块结构网格上的Kershaw扩散格式

Kershaw格式是在四边形结构网格上求解扩散方程的-种经典格式.基于对Kershaw格式中"流"的深入理解,将其拓展到包含非结构点的多块结构网格,分别推导退化非结构点和强化非结构点情况下的Kershaw格式,拓展的Kershaw格式满足流连续条件.三个数值算例的计算结果与精确解吻合得很好,表明将Kershaw格式拓展到多块结构网格的正确性和有效性.     

R?ssler混沌系统的脉冲同步

对R?ssler混沌系统给出了脉冲同步的方法．该方法基于脉冲微分方程理论,根据同步误差进行线性反馈,对系统施加脉冲作用,改变其状态变量,使两个系统同步．同时给出了一个同步的渐近稳定性条件,并进行数值仿真验证了该方法的正确性和抗噪声能力. 关键词：     

Networks, dynamics, and modularity

The identification of general principles relating structure to dynamics has been a major goal in the study of complex networks. We propose that the special case of linear network dynamics provides a natural framework within which a number of interesting yet tractable problems can be defined. We report the emergence of modularity and hierarchical organization in evolved networks supporting asymptotically stable linear dynamics. Numerical experiments demonstrate that linear stability benefits from the presence of a hierarchy of modules and that this architecture improves the robustness of network stability to random perturbations in network structure. This work illustrates an approach to network science which is simultaneously structural and dynamical in nature.     

基于粒子群优化的自适应模糊制冷控制算法

为解决空间斯特林制冷机和探测器热负载不确定及存在变化的问题,提出了自适应模糊PID制冷控制。在空间环境中使用的斯特林制冷机参数会随着时间的变化而发生改变,探测器负载也会随着工作模式和工作时间的变化而变化,整个制冷系统涉及的变量多,参数非线性。采用传统的控制方法,在固定的单一条件、环境下得到的控制参数,环境和负载发生变化后容易性能变差甚至不稳定,控制精度和稳定性不能满足使用要求。设计了一种自适应斯特林制冷机控制器,通过综合自适应模糊PID控制的方法,采用粒子群优化算法调整控制参数以减小代价函数。通过仿真和试验验证算法的有效性和鲁棒性。