Loading...
Association Journal of CSIAM
Supervised by Ministry of Education of PRC
Sponsored by Xi'an Jiaotong University
ISSN 1005-3085  CN 61-1269/O1

Table of Content

    15 February 2015, Volume 32 Issue 1    Next Issue
    Infrared Dim and Small Target Detection Based on SVD
    TIAN Chao, GAO Chen-qiang
    2015, 32 (1):  1-10.  doi: 10.3969/j.issn.1005-3085.2015.01.001
    Abstract ( 22 )   PDF (1050KB) ( 7 )   Save
    Considering the shortage of conventional single-image-SVD-based infrared dim and small target detection method, a new sequential-image-SVD-based method is proposed in this paper. Firstly, an image matrix is constructed upon sequential images and then decomposed by the SVD algorithm, and consequently the corresponding singular values and singular vectors are achieved. Secondly, new sequential feature images are reconstructed by using processed singular values and singular vectors. Thirdly, the selected feature images from sequential feature images are further processed to enhance the small targets and suppress the clutter background. Fourthly, the small targets are obtained by segmenting feature images. Finally, the detection task of small target detection is completed by refining the inter-frame and intra-frame target locations. Experimental results show that the proposed algorithm is robust and with real-time efficiency.
    Related Articles | Metrics
    Construction of Nonseparable Orthogonal Wavelet Tight Frames with Symmetry
    FENG Yan, SHEN Yan-feng, YANG Shou-zhi, YUAN De-hui
    2015, 32 (1):  11-20.  doi: 10.3969/j.issn.1005-3085.2015.01.002
    Abstract ( 17 )   PDF (809KB) ( 5 )   Save
    Nonseparable wavelets can capture high frequency information of multi-dimensional signal compared with separable ones, and orthogonal frames can avoid pre-filtering in processing of multi-dimensional signal. In order to process multi-dimensional signal efficiently, this paper studies the construction of nonseparable orthogonal wavelet tight frames constructed by unidimenstional one through tensor product, and multiplied by a trigonometric polynomial with some properties. Moreover, if the unidimenstional one is symmetric, the one so constructed is also symmetric. Finally, an example is provided.
    Related Articles | Metrics
    Global Optimality Conditions for Quadratic Program Problems with Quadratic Constraints
    ZHOU Xue-gang
    2015, 32 (1):  21-28.  doi: 10.3969/j.issn.1005-3085.2015.01.003
    Abstract ( 24 )   PDF (132KB) ( 7 )   Save
    In this paper, sufficient global optimality conditions are presented for nonconvex quadratic programming problems with quadratic constraints as well as hyperrectangle constr-aints. The new conditions are obtained by making use of quadratic underestimators of quadratic function. We first introduce how to construct quadratic underestimators of quadratic function. Then, by using convex quadratic underestimators of the Lagrangian function at the Karush-Kuhn-Tucker point, we establish sufficient global optimality conditions for nonconvex quadratic programming problems. And we propose sufficient global optimality conditions by utilizing the minimum eigenvalue and quadratic underestimators. Finally, by using quadratic underestimators, we establish the sufficient condition for nonconvex quadratic programming problems with quadratic constraints.
    Related Articles | Metrics
    On Newton-Triangle Splitting Methods for the Systems of Nonlinear Algebraic Equations
    HU Ji-yang, WANG Chuan-long, WEN Rui-ping
    2015, 32 (1):  29-38.  doi: 10.3969/j.issn.1005-3085.2015.01.004
    Abstract ( 24 )   PDF (167KB) ( 5 )   Save
    The Triangle Splitting iteration method is an effective iteration method for solving large-scale sparse non-Hermitian positive definite system of linear algebraic equations. By making use of the Triangle Splitting iteration method on non-Hermitian positive definite matrices as the inner solver of the inexact Newton method, we establish a class of inexact Newton-Triangle Splitting iteration methods for solving the large-scale sparse system of nonlinear algebraic equa-tions with positive definite Jacobian matrices in the paper. For this class of inexact Newton methods, two types of local convergence theorems are proved under proper conditions. The numerical results are given to examine their feasibility and effectiveness. The numerical implementations also show that the Newton-Triangle Splitting methods have advantages over Newton-BTSS methods with less computation time and iteration steps.
    Related Articles | Metrics
    Quadratic Inverse Eigenvalue Problem Under Double Complex Eigenvalues
    HUANG Xian-tong
    2015, 32 (1):  39-49.  doi: 10.3969/j.issn.1005-3085.2015.01.005
    Abstract ( 23 )   PDF (129KB) ( 5 )   Save
    The inverse eigenvalue problem is an important method for designing the circuit or the mass-spring system. The paper considers the following circuit design problem: given some information of the inductance matrix $M$, resistance matrix $C$, and capacitance matrix $K$, determine the rest data such that the system has the prescribed frequency. We transform the problem into a quadratic inverse eigenvalue problem under double complex eigenvalues. The existence and the expression of the solution are derived by solving the quadratic character determinant equations. We present the algorithm and numerical examples, which show that the obtained result is correct.
    Related Articles | Metrics
    Simulating Viscoelastic Flow Field Based on DCQ-QUICK Scheme
    ZHOU Wen, XIE Yan, OUYANG Jie, LI Qiang
    2015, 32 (1):  50-60.  doi: 10.3969/j.issn.1005-3085.2015.01.006
    Abstract ( 18 )   PDF (1218KB) ( 4 )   Save
    The discrete for the convection term is one of the main difficulties for the num-erical solution of viscoelastic fluid flow. In this paper, for the conservation equations and the Oldroyd-B constitutive equation, a deferred correction method coupled with high order Q-QUICK scheme for the computation of the convection flux is proposed. This method is designed based on the finite volume method on unstructured collocated grids. The planar poiseuille viscoelastic flow is simulated numerically to verify the high precision and stability of the proposed method. In the simulation of 4:1 contraction viscoelastic flow, the changes of the stream lines and stresses as well as growing of the salient corner vortex versus the Weissenberg numbers are revealed. The numerical results show that the numerical method is capable of expanding the range of the Weissenberg numbers for nonlinear viscoelastic fluid.
    Related Articles | Metrics
    A Note on Solutions to Systems of Fractional Differential Equations with Higher Order
    ZHENG Yan-ping, LI Sheng-li
    2015, 32 (1):  61-71.  doi: 10.3969/j.issn.1005-3085.2015.01.007
    Abstract ( 18 )   PDF (151KB) ( 4 )   Save
    Fractional-order derivatives provide a powerful instrument for describing the memory and hereditary of different substances. A growing number of works from various fields throughout science and engineering deal with dynamical systems expressed by fractional-order equations. In this paper, the existence and uniqueness of the solution to the initial problem for systems of fractional differential equations are derived. Based on the Schur theorem, the existence and uniqueness of the local solution to the linear system are given. Furthermore, the effectiveness of the proposed method is verified by a simulated example. For a nonlinear system, the existence and uniqueness of the solution are obtained by the Schauder fixed point theorem and the Banach fixed point theorem, respectively.
    Related Articles | Metrics
    A Predator-prey System with Time Delay and Stage Structure for the Predator and the Prey
    WANG Ling-shu, FENG Guang-hui
    2015, 32 (1):  72-84.  doi: 10.3969/j.issn.1005-3085.2015.01.008
    Abstract ( 19 )   PDF (165KB) ( 7 )   Save
    The stability of a predator-prey model with time delay and stage structure for both the predator and the prey is investigated. By analyzing the corresponding characteristic equations, sufficient conditions are given respectively for the local stability of each of feasible equilibria of the system and the existence of a Hopf bifurcation at the positive equilibrium. By means of the persistence theory on infinite dimensional systems, it is proven that the system is permanent if the positive equilibrium exists. By using the Lyapunov functions and the LaSalle invariant principle, sufficient conditions are derived respectively for the global stability of each of feasible equilibria of the model.
    Related Articles | Metrics
    The N-S Bifurcation of Feedback Control Model with Piecewise Constant Arguments and Interference
    CHEN Si-yang, JIN Bao
    2015, 32 (1):  85-97.  doi: 10.3969/j.issn.1005-3085.2015.01.009
    Abstract ( 19 )   PDF (487KB) ( 5 )   Save
    The dynamics of the feedback control model on a single population with piecewise constant arguments and interference are investigated in this paper. A difference model which can equi-valently describe the dynamical behavior of the original differential model is deduced. Based on the analysis of the eigenvalues and Schur-Cohn criterion, the sufficient conditions for local asymptotic stability of the positive equilibrium are achieved. Moreover, by choosing the intrinsic growth rate of the population as the bifurcation parameter and applying the bifurcation and center manifold theories, the existence conditions for the Neimark-Sacker bifurcation of this difference model is derived. Finally, some numerical examples substantiating our theoretical predictions are given and the numerical simulations also show that: 1) the dynamics of the single population of feedback control model are very complex when we consider piecewise constant arguments and interference; and 2) the positive equilibrium of the model switches from stable to unstable as the intrinsic growth rate of population increases beyond a critical value, at which the unique supercritical Neimark-Sacker bifurcation will occur.
    Related Articles | Metrics
    Probability Proof of Some Special Combination Sequence
    Alatantaogesi, Wuyungaowa
    2015, 32 (1):  98-106.  doi: 10.3969/j.issn.1005-3085.2015.01.010
    Abstract ( 21 )   PDF (122KB) ( 5 )   Save
    In many cases we can use combinatorial methods, e.g., the coefficient method and the Riordan arrays etc., to prove combinatorial identities. In this paper, by making use of some unusual techniques, we achieve some specific combinatorial identities. Specifically, we derive some new identities involving two kinds of Stirling numbers, reciprocal of binomial coefficient, harmonic numbers, Bell numbers and the number of derangements, by utilizing given probability expressions, properties of mathematical expectation, binomial identities and polynomial identities.
    Related Articles | Metrics
    Triangle Tower Network: a New Class of Interconnection Network
    SHI Hai-zhong, BAI Ya-lan, WANG Guo-liang, HU Yan-hong
    2015, 32 (1):  107-115.  doi: 10.3969/j.issn.1005-3085.2015.01.011
    Abstract ( 19 )   PDF (176KB) ( 6 )   Save
    In this paper, we propose and analyze a new interconnection network called triangle tower graph/network. It is maximally connected and tightly super-connected, for $n>4$ or $n=4$, i.e. the connectivity $\kappa(TT_{n})$ of $TT_{n}$ is $2n-3$. The star graph is a specific subgraph of the proposed triangle tower graph. Therefore, the triangle tower graph not only inherits many good capabilities possessed by the star graph (e.g., vertex symmetry, connectivity, vertex transition, etc.), but also shows that $S_{n}$ can be embedded into $TT_{n}$ with digit 1. The proposed triangle tower graph is superior to the traditional hypercube and bubble-sort graph with respect to diameter, connectivity and conditional vertex connectivity as that these three graphs have approximately similar numbers of vertices. The diameter and average distance are presented for the proposed network. We also propose one variety conjectures on Hamiltonicity of triangle tower graph and prove conjectures are true for $n=3,4$ and $n=5,6,~k=1,2$.
    Related Articles | Metrics
    Residual a Posteriori Error Estimate of Two Level Finite Element Method for Natural Convection Problem
    ZHANG Yun-zhang, HOU Yan-ren, WEI Hong-bo
    2015, 32 (1):  116-130.  doi: 10.3969/j.issn.1005-3085.2015.01.012
    Abstract ( 21 )   PDF (154KB) ( 9 )   Save
    This paper presents the a posteriori error estimate of residual for natural convection problem, which is computed by the two level Newton finite element method. The a posteriori error estimate contains additional terms in comparison to the one obtained by the standard one level finite element method. The action of the add-itional terms in the error estimate is investigated by studying their asymptotic behaviour. For optimally scaled meshes between coarse and fine meshes of the two level Newton finite element method, the additional terms are not of higher convergence order than the order of the numerical solution. Numerical experiments verify the obtained theory results.
    Related Articles | Metrics
    An Inexact Smoothing Newton Method for Solving Symmetric Cone Complementarity Problem
    RUI Shao-ping
    2015, 32 (1):  131-144.  doi: 10.3969/j.issn.1005-3085.2015.01.013
    Abstract ( 21 )   PDF (131KB) ( 6 )   Save
    In this paper, we propose an inexact smoothing method for solving symmetric cone complementarity problem based on a one-parametric class of smoothing function which contains the FB smoothing function and the CHKS smoothing function as special cases. At each iteration, we use the GMRES iterative solver to obtain an approximate solution to the generated smoothing Newton linear system. Under suitable conditions, we obtain global convergence and local superlinear convergence of the proposed algorithm. Numerical results indicate that the proposed algorithm is effective for large-scale problem.
    Related Articles | Metrics
    A Two-level Finite Difference Method for Burger's Equation
    Zulhumar Kadir, LI Ning, HUANG Peng-zhan, FENG Xin-long
    2015, 32 (1):  145-158.  doi: 10.3969/j.issn.1005-3085.2015.01.014
    Abstract ( 19 )   PDF (197KB) ( 6 )   Save
    In this paper, a two-level finite difference scheme is presented for the numerical approximation of Burger's equation. The full nonlinear problem is solved on a coarse grid of size $H$, and a linear problem is solved on a fine mesh with mesh size $h$. The new difference scheme, which is the implicit one with unconditional stability and easy computation. The method we study provides an approximate solution with nearly the same error as the usual one-level solution, which involves solving one large nonlinear problem on a fine mesh with mesh size $h$. Hence, our method is capable of significantly saving computational time.
    Related Articles | Metrics