On the convergence of a generalized Reduced gradient algorithm for nonlinear programming problems

Despite of their excellent numerical performance for solving practical nonlinear programming problems, the theoretical convergence behavior of generalized reduced gradient algorithms has been investigated very seldom in the literature. One specific class of generalized reduced gradient methods will be presented for which a global convergence result can be shown, i.e. the approximation of a Kuhn-Tucker point starting from arbitrary initial values. The relationship of the proposed variant with some other versions of generalized reduced gradient algorithms will be discussed.     

圆形区域分散布局问题研究

针对圆形区域分散布局问题, 文中给出了一个带约束的非线性规划模型. 当布局点数量较少时, 通过将模型转化为无约束优化问题, 利用梯度方法进行求解; 对于布局点数量较多的情况, 提出了一个界为1/2的多项式时间的近似算法, 并进行了相应的算例分析, 进一步来验证算法解的合理性. 研究的结论及方法一定程度上丰富和完善了圆形区域的分散布局理论.     

Superconvergent Cluster Recovery Method for the Crouzeix-Raviart Element

In this paper, we propose and numerically investigate a superconvergent cluster recovery (SCR) method for the Crouzeix-Raviart (CR) element. The proposed recovery method reconstructs a $C^0$ linear gradient. A linear polynomial approximation is obtained by a least square fitting to the CR element approximation at certain sample points, and then taken derivatives to obtain the recovered gradient. The SCR recovery operator is superconvergent on uniform mesh of four patterns. Numerical examples show that SCR can produce a superconvergent gradient approximation for the CR element, and provide an asymptotically exact error estimator in the adaptive CR finite element method.     

A combined direction stochastic approximation algorithm

The stochastic approximation problem is to find some root or minimum of a nonlinear function in the presence of noisy measurements. The classical algorithm for stochastic approximation problem is the Robbins-Monro (RM) algorithm, which uses the noisy negative gradient direction as the iterative direction. In order to accelerate the classical RM algorithm, this paper gives a new combined direction stochastic approximation algorithm which employs a weighted combination of the current noisy negative gradient and some former noisy negative gradient as iterative direction. Both the almost sure convergence and the asymptotic rate of convergence of the new algorithm are established. Numerical experiments show that the new algorithm outperforms the classical RM algorithm.     

Modeling and numerical simulation of linear and nonlinear spacecraft attitude dynamics and gravity gradient moments: A comparative study

In this paper linear and nonlinear models of spacecraft attitude dynamics equations and gravity gradient moments are investigated. In addition, effects of gravity gradient moments on attitude dynamics of the satellite are studied. The purpose of this paper is to present a comparison between nonlinear and linear models of spacecraft attitude dynamics and gravity gradient moments in order to determine divergence of linear approximation from the nonlinear model. Simulation results indicate that designer of spacecraft attitude control subsystem should be meticulous in applying linear approximation of equations especially in low earth orbits. Consequently, finding an upper bound for small angle to keep the linear model valid and precise enough would be a vital part of using linear approximation. Results supported by numerical examples demonstrate various features of this study.     

Meshfree Approximation for Stochastic Optimal Control Problems

In this work,we study the gradient projection method for solving a class of stochastic control problems by using a mesh free approximation ap-proach to implement spatial dimension approximation.Our main contribu-tion is to extend the existing gradient projection method to moderate high-dimensional space.The moving least square method and the general radial basis function interpolation method are introduced as showcase methods to demonstrate our computational framework,and rigorous numerical analysis is provided to prove the convergence of our meshfree approximation approach.We also present several numerical experiments to validate the theoretical re-sults of our approach and demonstrate the performance meshfree approxima-tion in solving stochastic optimal control problems.     

Pointwise superconvergence of the gradient for the linear tetrahedral element

We consider the finite element approximation to the solution of a self-adjoint, second-order elliptic boundary value problem in three dimensions over a fully uniform mesh of piecewise linear tetrahedral elements. Although the resulting approximation to the gradient is optimal for functions from the approximating space, it is, however, only O(h). We show how this can be improved by the recovery, from the finite element solution, of an approximation to the gradient, which is pointwise of a higher order of accuracy than that of the gradient of the finite element approximation. This approximation, termed a recovered gradient function, is, thus, superconvergent. The major task of our analysis is the establishing of an (almost) constant bound on the W seminorm of the finite element approximation to a smoothed derivative Green's function. © 1994 John Wiley & Sons, Inc.     

Approximating Gradients with Continuous Piecewise Polynomial Functions

Motivated by conforming finite element methods for elliptic problems of second order, we analyze the approximation of the gradient of a target function by continuous piecewise polynomial functions over a simplicial mesh. The main result is that the global best approximation error is equivalent to an appropriate sum in terms of the local best approximation errors on elements. Thus, requiring continuity does not downgrade local approximation capability and discontinuous piecewise polynomials essentially do not offer additional approximation power, even for a fixed mesh. This result implies error bounds in terms of piecewise regularity over the whole admissible smoothness range. Moreover, it allows for simple local error functionals in adaptive tree approximation of gradients.     

分片C~2凸函数Moreau-Yosida逼近的分片光滑性质

对分片 Ｃ２凸函数的 Ｍｏｒｅａｕ－Ｙｏｓｉｄａ逼近研究了它的梯度性质,引进了序列常秩约束条件,在此条件下证明了梯度函数具有分片光滑性质．     

非奇异阵特征极值和条件数的近似计算

本文从共轭梯度法的公式推导出对称正定阵A与三对角阵B的相似关系,B的元素由共轭梯度法的迭代参数确定.因此,对称正定阵的条件数计算可以化成三对角阵条件数的计算,并且可以在共轭梯度法的计算中顺带完成.它只需增加O(s)次的计算量,s为迭代次数.这与共轭梯度法的计算量相比是可以忽略的.当A为非对称正定阵时,只要A非奇异,即可用共轭梯度法计算ATA的特征极值和条件数,从而得出A的条件数.对不同算例的计算表明,这是一种快速有效的简便方法.     

Star-shaped approximation approach for stochastic programming problems with probability function

Star-shaped probability function approximation is suggested. Conditions of log-concavity and differentiability of approximation function are obtained. The method for constructing stochastic estimates of approximation function gradient and stochastic quasi-gradient algorithm for probability function maximization are described in the paper     

Line search termination criteria for collinear scaling algorithms for minimizing a class of convex functions

Summary. Many successful quasi-Newton methods for optimization are based on positive definite local quadratic approximations to the objective function that interpolate the values of the gradient at the current and new iterates. Line search termination criteria used in such quasi-Newton methods usually possess two important properties. First, they guarantee the existence of such a local quadratic approximation. Second, under suitable conditions, they allow one to prove that the limit of the component of the gradient in the normalized search direction is zero. This is usually an intermediate result in proving convergence. Collinear scaling algorithms proposed initially by Davidon in 1980 are natural extensions of quasi-Newton methods in the sense that they are based on normal conic local approximations that extend positive definite local quadratic approximations, and that they interpolate values of both the gradient and the function at the current and new iterates. Line search termination criteria that guarantee the existence of such a normal conic local approximation, which also allow one to prove that the component of the gradient in the normalized search direction tends to zero, are not known. In this paper, we propose such line search termination criteria for an important special case where the function being minimized belongs to a certain class of convex functions. Received February 1, 1997 / Revised version received September 8, 1997     

A posteriori FE error control for p-Laplacian by gradient recovery in quasi-norm

A posteriori error estimators based on quasi-norm gradient recovery are established for the finite element approximation of the p-Laplacian on unstructured meshes. The new a posteriori error estimators provide both upper and lower bounds in the quasi-norm for the discretization error. The main tools for the proofs of reliability are approximation error estimates for a local approximation operator in the quasi-norm.

     

An entropic gradient structure for quasi-steady-state approximations of chemical reactions

Passing to the limit of an infinite reaction rate in a slow-fast system of chemical reactions provides a quasi-steady state approximation (QSSA) of these systems. In case of reactions with detailed balance condition, this approximation includes a dimension reduction to a smaller state space. We show that the limit dynamics carry an entropic gradient structure on this smaller space. (© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

On a global superconvergence of the gradient of linear triangular elements

We study a simple superconvergent scheme which recovers the gradient when solving a second-order elliptic problem in the plane by the usual linear elements. The recovered gradient globally approximates the true gradient even by one order of accuracy higher in the L²-norm than the piecewise constant gradient of the Ritz—Galerkin solution. A superconvergent approximation to the boundary flux is presented as well.     

The Gradient Superconvergence of Bilinear Finite Volume Element for Elliptic Problems

We study the gradient superconvergence of bilinear finite volume element (FVE) solving the elliptic problems. First, a superclose weak estimate is established for the bilinear form of the FVE method. Then, we prove that the gradient approximation of the FVE solution has the superconvergence property: where denotes the average gradient on elements containing point $P$ and $S$ is the set of optimal stress points composed of the mesh points, the midpoints of edges and the centers of elements.     

Numerical Study of Hydrothermal Wave Suppression in Thermocapillary Flow Using a Predictive Control Method

Hydrothermal waves in flows driven by thermocapillary and buoyancy effects are suppressed by applying a predictive control method. Hydrothermal waves arise in the manufacturing of crystals, including the “open boat” crystal growth process, and lead to undesirable impurities in crystals. The open boat process is modeled using the two-dimensional unsteady incompressible Navier–Stokes equations under the Boussinesq approximation and the linear approximation of the surface thermocapillary force. The flow is controlled by a spatially and temporally varying heat flux density through the free surface. The heat flux density is determined by a conjugate gradient optimization algorithm. The gradient of the objective function with respect to the heat flux density is found by solving adjoint equations derived from the Navier–Stokes ones in the Boussinesq approximation. Special attention is given to heat flux density distributions over small free-surface areas and to the maximum admissible heat flux density.     

A New Approximation of the Matrix Rank Function and Its Application to Matrix Rank Minimization

The matrix rank minimization problem is widely applied in many fields such as control, signal processing and system identification. However, the problem is NP-hard in general and is computationally hard to directly solve in practice. In this paper, we provide a new approximation function of the matrix rank function, and the corresponding approximation problems can be used to approximate the matrix rank minimization problem within any level of accuracy. Furthermore, the successive projected gradient method, which is designed based on the monotonicity and the Fréchet derivative of these new approximation function, can be used to solve the matrix rank minimization this problem by using the projected gradient method to find the stationary points of a series of approximation problems. Finally, the convergence analysis and the preliminary numerical results are given.     

AN EXPANDED CHARACTERISTIC-MIXED FINITE ELEMENT METHOD FOR A CONVECTION-DOMINATED TRANSPORT PROBLEM

In this paper, we propose an Expanded Characteristic-mixed Finite Element Method for approximating the solution to a convection dominated transport problem. The method is a combination of characteristic approximation to handle the convection part in time and an expanded mixed finite element spatial approximation to deal with the diffusion part. The scheme is stable since fluid is transported along the approximate characteristics on the discrete level. At the same time it expands the standard mixed finite element method in the sense that three variables are explicitly treated： the scalar unknown, its gradient, and its flux. Our analysis shows the method approximates the scalar unknown, its gradient, and its flux optimally and simultaneously. We also show this scheme has much smaller time-truncation errors than those of standard methods. A numerical example is presented to show that the scheme is of high performance.     

Shape sensitivity analysis of an elastic contact problem: Convergence of the Nitsche based finite element approximation

In a recent work, we introduced a finite element approximation for the shape optimization of an elastic structure in sliding contact with a rigid foundation where the contact condition (Signorini’s condition) is approximated by Nitsche’s method and the shape gradient is obtained via the adjoint state method. The motivation of this work is to propose an a priori convergence analysis of the numerical approximation of the variables of the shape gradient (displacement and adjoint state) and to show some numerical results in agreement with the theoretical ones. The main difficulty comes from the non-differentiability of the contact condition in the classical sense which requires the notion of conical differentiability.