基于小波神经网络方法的桥梁结构损伤识别研究

桥梁结构在服役期间会承受复杂的荷载,长期使用会不可避免地出现各种损伤.若这些损伤不能被及时发现和适当处理,将有可能造成严重的事故.因此,桥梁结构的局部小损伤识别对于其及时检修有重要意义.通常,损伤结构的全局动态特性测试可能对局部的结构损伤不敏感,特别是对小损伤,这就需要从结构动态响应信号中提取对损伤更敏感的特征量.建立了桥梁结构的有限元模型并进行动力特性分析;采用小波包分析方法处理结构动态响应信号以构造结构损伤指标,并结合结构损伤指标和人工神经网络方法进行桥梁结构的损伤定位.     

基于一种灰色综合关联模型的区域客运结构优化研究

为实现客运结构总体区域优化,运用灰色系统理论中的灰色综合关联分析模型,以珠三角地区客运结构为研究对象进行实证研究,建立了各种客运方式的旅客周转量与相关经济社会主要指标间的灰色综合关联矩阵,对客运结构系统特征及相关影响因素作优势分析.研究结论表明:城市化程度对客运结构影响最大;其次是人均可支配收入,再次是人均生产总值;人口总数对客运结构影响小,但对客运总量影响大.珠三角地区在中短途客运方面,应大力发展城市轨道交通(城际铁路和地下铁道);在长途客运方面,应大力发展高速铁路和民航.     

我国国债期限结构特征研究

论文对1981年以来我国国债发行市场和交易市场期限结构的特征进行了统计分析,并绘制了不同时期的期限结构曲线.论文最后采用纯预期理论对我国期限结构曲线特征进行了研究.     

无穷小量等价类集的代数结构

实现了无穷小量集合的分类,建立了两种无穷小等价类的分划,指出了无穷小等价类的存在性与元素结构,为寻求无穷小等价类的元素特征提供了理论基础.并在此基础之上,分析了无穷小等价类的代数结构,为进一步运用无穷小量进行各种运算提供了系统的理论基础.     

A Study on the Fatigue Life of the Laser Additive Manufactured Metallic Material北大核心CSCD

随着军用飞机上越来越多的结构件使用增材制造（AM）技术成形,对增材制造材料和结构疲劳特性的研究就变得十分迫切.为了研究激光增材制造（选区激光熔化成形,SLM）铝合金以及钛合金的疲劳寿命特性,设计了一系列带各种结构细节的模拟试验件,进行常幅谱和随机谱下的试验,统计分析了各组试验的基本可靠性寿命,并得到了可靠度安全寿命曲线.断口分析发现铝合金试验件缺陷较多,存在混合失效特征,而钛合金试验件疲劳分散性较锻件大.     

基于SCN函数共轭梯度方向的稀疏向量特征提取算法

稀疏向量特征提取是指在优化时利用各种范数对解进行约束,从而获得带有稀疏特征的最优解,其广泛应用于复杂系统中的机器学习、深度学习和大数据分析等领域的特征提取问题.大量的研究表明各种范数如L₀范数、L₁范数和L₂范数的方法都存在各自的缺点,主要表现在越容易求解的范数越不精准稀疏,越精准稀疏的范数越难求解.文章提出了一种基于SCN函数共轭梯度方向的稀疏向量特征发现算法(CGDL),稀疏向量特征发现可以用一个稀疏特征提取优化模型建立,其目标函数是一个SCN函数,对其中的L₀范数进行转换,形成一个具有特殊结构优化问题,这个问题等价于双层规划的凸-凹极小极大化问题,这类问题可以解决稀疏回归、图像特征和压缩感知等问题.文章给出了上述模型的稀疏特征提取算法的详细计算步骤和收敛性分析证明,并且对给定的实际数据集和高维模拟数据集对算法的有效性、复杂性和收敛速度进行了数值对比实验,表明了该算法在精准度和稀疏性上显著优于其他对比方法,并且具有较好的收敛速度.     

基于多尺度局部结构约束的特征点匹配算法

对于含有重复模式较多、低重叠区域的图像,特征匹配时会存在离群点较多及伪同构等问题,从而影响匹配的精度.为了提高匹配算法的精度和鲁棒性,本文提出了一种基于多尺度局部结构相似性的特征点匹配算法(MLSS).首先定义局部结构相似性度量方法,即多尺度近邻结构相似性及多尺度局部拓扑结构相似性.基于局部结构相似性,构造特征匹配模型,并采用确定性退火思想去除离群点,实现特征点集的精确和快速的匹配.使用Daisy数据集、VGG数据集和航空遥感数据集中的22组典型数据对算法的性能进行验证,匹配结果显示,该算法的精度和鲁棒性优于其他四种经典的特征匹配方法.     

幂环的结构与分类

完整地研究了各种幂环的结构及它们之间的关系. 对幂环进行了分类, 并构造了各类的子环列和理想列.     

Witt代数和Virasoro代数上的Poisson代数结构

Poisson代数是指同时具有结合代数结构和李代数结构的一类代数,其结合代数结构和李代数结构满足Leibniz法则.确定了特征为0和特征为p>0的基域上的Witt代数和Virasoro代数上的Poisson代数结构.     

分层广义线性模型在准备金评估中的建模研究

考虑到赔付流量三角形数据同一事故年反复观测的纵向特征以及数据结构的层次性,建立了分层广义线性模型.与通常的随机模型相比,分层广义线性模型不但可以选择条件反应变量的分布而且风险参数分布范围也更加广泛.利用h-似然函数估计分层广义线性模型的模型参数,降低了计算量.为使模型具有可比性,评估模型的预测精度,推导了模型预测误差的估计式.为充分利用已知赔付信息,将赔付额和赔付次数两种赔付信息纳入未决赔款准备金评估模型,建立了两阶段分层广义线性模型.在线性预测量中考虑了各种固定效应和随机效应以及模型结构的散布参数,改进了线性预估量结构.研究表明:分层广义线性模型对于数据的各种分布及形式都具有很好的适应性,更加符合保险实务现实的赔付规律.     

Normalization of sunspot cycles and eigen mode analysis

The smoothed monthly sunspot numbers of the previous 22 complete sunspot cycles are normalized in time domain, and then an eigen mode analysis is carried out to draw the principle factors (or components) in the cycles. The results show that the main characteristics of the solar cycles can be described fairly well by the first 5 eigen modes. The obtained eigen modes are used to predict the declining phase of cycle 23 on the basis of the data prior to its maximum. The prediction indicates that cycle 23 will last for 127 months to December 2006, with the minimum of 6.2.     

Normalization of sunspot cycles and eigen mode analysis

The smoothed monthly sunspot numbers of the previous 22 complete sunspot cycles are normalized in time domain, and then an eigen mode analysis is carried out to draw the principle factors (or components) in the cycles. The results show that the main characteristics of the solar cycles can be described fairly well by the first 5 eigen modes. The obtained eigen modes are used to predict the declining phase of cycle 23 on the basis of the data prior to its maximum. The prediction indicates that cycle 23 will last for 127 months to December 2006, with the minimum of 6.2.     

Eigen‐frequencies in thin elastic 3‐D domains and Reissner–Mindlin plate models

The eigen‐frequencies of elastic three‐dimensional thin plates are addressed and compared to the eigen‐frequencies of two‐dimensional Reissner–Mindlin plate models obtained by dimension reduction. The qualitative mathematical analysis is supported by quantitative numerical data obtained by the p‐version finite element method. The mathematical analysis establishes an asymptotic expansion for the eigen‐frequencies in power series of the thickness parameter. Such results are new for orthotropic materials and for the Reissner–Mindlin model. The 3‐D and R–M asymptotics have a common first term but differ in their second terms. Numerical experiments for clamped plates show that for isotropic materials and relatively thin plates the Reissner–Mindlin eigen‐frequencies provide a good approximation to the three‐dimensional eigen‐frequencies. However, for some anisotropic materials this is no longer the case, and relative errors of the order of 30 per cent are obtained even for relatively thin plates. Moreover, we showed that no shear correction factor is known to be optimal in the sense that it provides the best approximation of the R–M eigen‐frequencies to their 3‐D counterparts uniformly (for all relevant thicknesses range). Copyright © 2002 John Wiley & Sons, Ltd.     

Composition of binary integral quadratic forms via integral 2 × 2 matrices and composition of matrix classes

A treatment by integral matrices is given for composition of pairs of integral quadratic forms with the same discriminant. The forms are associated with a pair of similar 2 × 2 matrices AB with irrational eigen values which generate the maximal order. The most general integral similarity between AB is given by a matrix whose entries are linear forms in two indeterminates with integral coefficients. This matrix is a "compositum" of two factors of the same nature. By equating determinants a composition of two quadratic forms results. The method can be generalized to n × n matrices.     

Extended Center Manifold, Global Bifurcation and Approximate Solutions of Chen chaotic dynamical system

The study of the chaotic Chen dynamic System (CDS) has been a recent focus in the literature, with numerous works exploring its various chaotic features. However, the majority of these studies have relied primarily on numerical techniques to investigate nonlinear dynamic systems (NLDSs). In this context, our aim is to derive approximate analytical solutions for the CDS by developing an iterative scheme. We have proven the convergence theorem for this scheme, which ensures that our iterative process will converge to the exact solution. Additionally, we introduce a new method for constructing the extended center manifold, a critical component in the analysis of dynamical systems. The characteristics of the global bifurcation of the system components within the parameter space are explored. The error analysis of the iterated solutions demonstrates the efficiency of the present technique. We present both three-dimensional (3D) and two-dimensional (2D) phase portraits of the system. The 3D portrait reveals a feedback loop pattern, while the 2D portrait, which represents the interaction of the system components, exhibits multiple pools and cross pools. Furthermore, we illustrate the global bifurcation by visualizing the components of the CDS against the space parameters. The sensitivity of CDS to innitesimal variations in the initial conditions (ICs) are tested. It is found that even minor changes can lead to signicant alterations in the system.     

广义Jacobi矩阵的逆特征问题

本文主要讨论广义Jacobi阵及多个特征对的广义Jacobi阵逆特征问题.通过相似变换将广义Jacobi阵变换为三对角对称矩阵,其特征不变、特征向量只作线性变换,再应用前人理论求得广义Jacobi阵元素ai,|bi|,|ci|有唯一解的充要条件及其具体表达式.     

3D-Hit: fast structural comparison of proteins on multicore architectures

3D-Hit is a well established method for rapid detection of structural similarities between proteins, which is widely used in various bioinformatics web servers (MetaServer, GRDB, 3D-Fun, Rosetta, etc.). The algorithm decomposes proteins into set of overlaping segments of 9–13 residues, then tries to match them using root mean square distance metric. The best aligned pairs of segments are selected as seeds for futher analysis. Those initial hits are expanded by iterative process in order to construct the global structural alignment by concatenating pairs of matching segments. The method has the same accuracy as the other state-of-the-art structural comparison algorithms (LGscore2, DALI), yet it provides much faster processing times, and can be used in a high-throughput setup as the structural module of bioinformatics pipelines. The method is optimized in terms of speed and accuracy to work on novel computer architectures, such as PowerXCell8i and Sun Constellation System. Here, we provide the source code of the 3D-Hit program, describe selected architectures on which the software was ported, present programing models, point out significant porting steps and sumarize performance comparisons.     

Dynamic programming approximation algorithms for the capacitated lot-sizing problem

This paper provides a new idea for approximating the inventory cost function to be used in a truncated dynamic program for solving the capacitated lot-sizing problem. The proposed method combines dynamic programming with regression, data fitting, and approximation techniques to estimate the inventory cost function at each stage of the dynamic program. The effectiveness of the proposed method is analyzed on various types of the capacitated lot-sizing problem instances with different cost and capacity characteristics. Computational results show that approximation approaches could significantly decrease the computational time required by the dynamic program and the integer program for solving different types of the capacitated lot-sizing problem instances. Furthermore, in most cases, the proposed approximate dynamic programming approaches can accurately capture the optimal solution of the problem with consistent computational performance over different instances.     

Computer system for kinematic and dynamic analysis synthesis of rigid and flexible multibody systems

In the paper an overview of a general numerical algorithm and program system library for deriving the kinematic constraint equations and dynamic equations of motion, as well as, computation of their first and second order partial derivatives with respect to kinematic parameters of motion, design parameters and mass and inertia characteristics for rigid and flexible multibody systems is presented. These are the main basic computational modules for implementation of kinematic and dynamic synthesis, optimization and design. The main theoretical basis consists in matrix methods for deriving the kinematic constraints and dynamic equations, as well as, the generalized Newton – Euler dynamic equations for rigid and flexible bodies, and finite element discretization in relative coordinates. Block–scheme of the computational procedures and problem oriented program compilation is presented. An example of kinematic synthesis of six–link path generating mechanism with singular points is presented. (© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)     

集中阻尼弦本征解的性质

利用Dirac δ函数,在全域建立并求解集中阻尼弦的动力学方程,导出其本征方程组、频率方程和本征函数的一般形式,推导了单项阻尼下本征函数的具体形式,并分析了中点阻尼对本征解的影响.同时,讨论了混合动力学系统在频率 阻尼关系、衰减率和完全抑制振动的最优阻尼3个方面既不同于连续系统,又不同于离散系统的特性:1）系统频率与其阻尼无关;2）各阶本征函数在单位时间内的衰减率都相同,衰减率与本征值的阶次无关;3）当阻尼取2时,系统衰减率趋于无穷大,系统不能发生任何有阻尼振动.