凸约束伪单调方程组的无导数投影算法

基于HS共轭梯度法的结构,本文在弱假设条件下建立了一种求解凸约束伪单调方程组问题的迭代投影算法.该算法不需要利用方程组的任何梯度或Jacobian矩阵信息,因此它适合求解大规模问题.算法在每一次迭代中都能产生充分下降方向,且不依赖于任何线搜索条件.特别是,我们在不需要假设方程组满足Lipschitz条件下建立了算法的全局收敛性和R-线收敛速度.数值结果表明,该算法对于给定的大规模方程组问题是稳定和有效的.     

Analysis of third-order methods for secular equations

Third-order numerical methods are analyzed for secular equations. These equations arise in several matrix problems and numerical linear algebra applications. A closer look at an existing method shows that it can be considered as a classical method for an equivalent problem. This not only leads to other third-order methods, it also provides the means for a unifying convergence analysis of these methods and for their comparisons. Finally, we consider approximated versions of the aforementioned methods.

     

求解简单界约束优化问题的一种逐次逼近法

１引言考虑变量带简单界约束的非线性规划问题：其中二阶连续可微,ａ＝（ａ１,ａ２,…,ａｎ）,ｂ＝（ｂ１,ｂ２,…,ｂｎ）,＋ｉ＝１,２,…,ｎ．问题（１）不仅是实际应用中出现的简单界约束最优化问题,而且相当一部分最优化问题可以把变量限制在有意义的区间内（参见［１］）．因此无论在理论方面还是在实际应用方面,都有研究此类问题并给出简便而有效算法的必要．假设ｆ是凸函数,记ｇ（ｘ）＝ｆ（ｘ）,则由Ｋ－Ｔ条件,问题（１）可化为求解下面的非光滑方程组：显然,（２）等价于易证,（３）等价于求解下面的非光滑方程…     

求解线性互补问题的一种新算法

本文针对线性互补问题,提出了与其等价的非光滑方程的逐次逼近阻尼牛顿法,并在一定条件下证明了该算法具有的全局收敛性．同时给出了一些数值例子,得到很好的数值结果．     

解可分离结构变分不等式的一种新的交替方向法

交替方向法是求解可分离结构变分不等式问题的经典方法之一, 它将一个大型的变分不等式问题分解成若干个小规模的变分不等式问题进行迭代求解. 但每步迭代过程中求解的子问题仍然摆脱不了求解变分不等式子问题的瓶颈. 从数值计算上来说, 求解一个变分不等式并不是一件容易的事情.因此, 本文提出一种新的交替方向法, 每步迭代只需要求解一个变分不等式子问题和一个强单调的非线性方程组子问题. 相对变分不等式问题而言, 我们更容易、且有更多的有效算法求解一个非线性方程组问题. 在与经典的交替方向法相同的假设条件下, 我们证明了新算法的全局收敛性. 进一步的数值试验也验证了新算法的有效性.     

An Inexact PRP Conjugate Gradient Method for Symmetric Nonlinear Equations

In this article, without computing exact gradient and Jacobian, we proposed a derivative-free Polak-Ribière-Polyak (PRP) method for solving nonlinear equations whose Jacobian is symmetric. This method is a generalization of the classical PRP method for unconstrained optimization problems. By utilizing the symmetric structure of the system sufficiently, we prove global convergence of the proposed method with some backtracking type line search under suitable assumptions. Moreover, we extend the proposed method to nonsmooth equations by adopting the smoothing technique. We also report some numerical results to show its efficiency.     

A class of conjugate gradient methods for convex constrained monotone equations

The recent designed non-linear conjugate gradient method of Dai and Kou [SIAM J Optim. 2013;23:296–320] is very efficient currently in solving large-scale unconstrained minimization problems due to its simpler iterative form, lower storage requirement and its closeness to the scaled memoryless BFGS method. Just because of these attractive properties, this method was extended successfully to solve higher dimensional symmetric non-linear equations in recent years. Nevertheless, its numerical performance in solving convex constrained monotone equations has never been explored. In this paper, combining with the projection method of Solodov and Svaiter, we develop a family of non-linear conjugate gradient methods for convex constrained monotone equations. The proposed methods do not require the Jacobian information of equations, and even they do not store any matrix in each iteration. They are potential to solve non-smooth problems with higher dimensions. We prove the global convergence of the class of the proposed methods and establish its R-linear convergence rate under some reasonable conditions. Finally, we also do some numerical experiments to show that the proposed methods are efficient and promising.     

A Superlinearly Convergent SSLE Algorithm for Optimization Problems with Linear Complementarity Constraints

In this paper we study a special kind of optimization problems with linear complementarity constraints. First, by a generalized complementarity function and perturbed technique, the discussed problem is transformed into a family of general nonlinear optimization problems containing parameters. And then, using a special penalty function as a merit function, we establish a sequential systems of linear equations (SSLE) algorithm. Three systems of equations solved at each iteration have the same coefficients. Under some suitable conditions, the algorithm is proved to possess not only global convergence, but also strong and superlinear convergence. At the end of the paper, some preliminary numerical experiments are reported.     

INEXACT DAMPED NEWTON METHOD FOR NONLINEAR COMPLEMENTARITY PROBLEMS

In this paper, we propose an inexact clamped Newton method for solving nonlinear complementarity problems based on the equivalent B-differentiable equations.Global convergence and locally quadratic convergence are obtained,and numerical results are given.     

Numerical method of lines for parabolic functional differential equations

Initial boundary value problems for nonlinear parabolic functional differential equations are transformed by discretization in space variables into systems of ordinary functional differential equations. A comparison theorem for differential difference inequalities is proved. Sufficient conditions for the convergence of the numerical method of lines are given. An explicit Euler method is proposed for the numerical solution of systems thus obtained. This leads to difference scheme for the original problem. A complete convergence analysis for the method is given.     

具有全局收敛性的求解对称非线性方程组的一个修改的信赖域方法

本文给出一个求解非线性对称方程组问题的修改的信赖域方法,在适当的条件下我们将建立此方法的全局收敛性.对给定的问题而言,数值结果表明此方法是有效的.     

ω-条件下Ulm-type方法的局部收敛性

由于牛顿法具有重要的理论基础和广泛的应用背景,它的收敛性得到了广泛研究([2,3,4,13,20,21,22,23]).—般而言,牛顿法的收敛性可以分成三类.一类是局部收敛性:已知方程(1)的解存在,初始点x0在该解的某个领域内时讨论牛顿法的收敛性([21,22,23]).     

Convergence of the Tamed Euler Method for Stochastic Differential Equations with Piecewise Continuous Arguments Under Non-global Lipschitz Continuous Coefficients

In this paper, we deal with the strong convergence of numerical methods for stochastic differential equations with piecewise continuous arguments (SEPCAs) with at most polynomially growing drift coefficients and global Lipschitz continuous diffusion coefficients. An explicit and time-saving tamed Euler method is used to solve this type of SEPCAs. We show that the tamed Euler method is bounded in pth moment. And then the convergence of the tamed Euler method is proved. Moreover, the convergence order is one-half. Several numerical simulations are shown to verify the convergence of this method.     

Local convergence theorems of Newton's method for nonlinear equations using outer or generalized inverses

We provide local convergence theorems for Newton's method in Banach space using outer or generalized inverses. In contrast to earlier results we use hypotheses on the second instead of the first Fréchet-derivative. This way our convergence balls differ from earlier ones. In fact we show that with a simple numerical example that our convergence ball contains earlier ones. This way we have a wider choice of initial guesses than before. Our results can be used to solve undetermined systems, nonlinear least squares problems and ill-posed nonlinear operator equations.     

Hybrid Newton-Type Method for a Class of Semismooth Equations

In this paper, we present a hybrid method for the solution of a class of composite semismooth equations encountered frequently in applications. The method is obtained by combining a generalized finite-difference Newton method to an inexpensive direct search method. We prove that, under standard assumptions, the method is globally convergent with a local rate of convergence which is superlinear or quadratic. We report also several numerical results obtained applying the method to suitable reformulations of well-known nonlinear complementarity problems.     

A projection method for a system of nonlinear monotone equations with convex constraints

In this paper, we propose a projection method for solving a system of nonlinear monotone equations with convex constraints. Under standard assumptions, we show the global convergence and the linear convergence rate of the proposed algorithm. Preliminary numerical experiments show that this method is efficient and promising. This work was supported by the Postdoctoral Fellowship of The Hong Kong Polytechnic University, the NSF of Shandong China (Y2003A02).     

A Globally Convergent Trust-Region Method for Large-Scale Symmetric Nonlinear Systems

This study presents a novel adaptive trust-region method for solving symmetric nonlinear systems of equations. The new method uses a derivative-free quasi-Newton formula in place of the exact Jacobian. The global convergence and local quadratic convergence of the new method are established without the nondegeneracy assumption of the exact Jacobian. Using the compact limited memory BFGS, we adapt a version of the new method for solving large-scale problems and develop the dogleg scheme for solving the associated trust-region subproblems. The sufficient decrease condition for the adapted dogleg scheme is established. While the efficiency of the present trust-region approach can be improved by using adaptive radius techniques, utilizing the compact limited memory BFGS adjusts this approach to handle large-scale symmetric nonlinear systems of equations. Preliminary numerical results for both medium- and large-scale problems are reported.     

An algorithm for a class of nonlinear complementarity problems with non-Lipschitzian functions

In this paper, we focus on solving a class of nonlinear complementarity problems with non-Lipschitzian functions. We first introduce a generalized class of smoothing functions for the plus function. By combining it with Robinson's normal equation, we reformulate the complementarity problem as a family of parameterized smoothing equations. Then, a smoothing Newton method combined with a new nonmonotone line search scheme is employed to compute a solution of the smoothing equations. The global and local superlinear convergence of the proposed method is proved under mild assumptions. Preliminary numerical results obtained applying the proposed approach to nonlinear complementarity problems arising in free boundary problems are reported. They show that the smoothing function and the nonmonotone line search scheme proposed in this paper are effective.     

混合互补问题的逐次逼近算法

针对混合互补问题 ,提出了与其等价的非光滑方程的逐次逼近算法 ,并在一定条件下证明了该算法的全局收敛性 .数值例子表明这一算法是有效的     

A variant of Steffensen’s method of fourth-order convergence and its applications

In this paper, a variant of Steffensen’s method of fourth-order convergence for solving nonlinear equations is suggested. Its error equation and asymptotic convergence constant are proven theoretically and demonstrated numerically. The derivative-free method only uses three evaluations of the function per iteration to achieve fourth-order convergence. Its applications on systems of nonlinear equations and boundary-value problems of nonlinear ODEs are showed as well in the numerical examples.