一类非线性规划问题的信赖域内点算法

本文对约束为线性的一类非线性优化问题提出了一种依赖域内点算法的,其中约束非负性要求一个仿射变换阵实现,其子问题变成了与个带仿射变换的线性等式约束的求解,我们证明了算法的有效性,在一定条件下证明了由算法产生的序列收敛到优化总理２的一阶稳定,点。     

不等式约束优化一个可行序列线性方程组算法

提出了求解非线性不等式约束优化问题的一个可行序列线性方程组算法. 在每次迭代中, 可行下降方向通过求解两个线性方程组产生, 系数矩阵具有较好的稀疏性. 在较为温和的条件下, 算法具有全局收敛性和强收敛性, 数值试验表明算法是有效的.     

一个求解P0函数非线性互补问题的非内部连续化算法

基于黄正海等2001年提出的光滑函数,本文给出一个求解P0函数非线性互补问题的非内部连续化算法.所给算法拥有一些好的特性.在较弱的条件下,证明了所给算法或者是全局线性收敛,或者是全局和局部超线性收敛.给出了所给算法求解两个标准测试问题的数值试验结果.     

一种求解不等式约束优化问题的光滑化算法

利用光滑函数建立了不等式约束优化问题KT条件的一个扰动方程组,提出了一个新的内点型算法.该算法在有限步终止时当前迭代点即为优化问题的一个精确稳定点.在一定条件下算法具有全局收敛性,数值试验表明该算法是有效的.     

求解不等式约束优化问题无严格互补松弛条件的QP-Free新算法

本文针对不等式约束优化问题,结合Facchinei-Fischer-Kanzow精确有效集识别技术,给出—个新的线性方程组与辅助方向相结合的可行下降算法.算法每步迭代只需求解一个降维的线性方程组或计算一次辅助方向,且获取辅助方向的投影矩阵只涉及近似有效约束集中的元素,问题规模大为减少,且当迭代次数充分大时,只需求解一个降维的线性方程组.无需严格互补松弛条件,算法全局且一步超线性收敛.     

一个具有线性约束的Goldfarb方法的改进算法

本文对具有线性约束的非线性规划问题给出一个Goldfarb方法的改进算法,并且在与[1]同样的条件下,给出了算法之超线性收敛性证明.     

一个结构简单的约束变尺度算法

本文我们考虑具有线性约束凹函数的最优化问题，利用我们的算法和变尺度修正公式，提出了一个结构简单的组合算法，并在「２」，「３」和「４」同样的假设条件下，证明了该算法的收敛性和超线性收敛速度，从而使该算法比原有各算法更具实用性。     

关于不等式约束的信赖域算法

对于具有不等式约束的非线性优化问题，本文给出一个依赖域算法，由于算法中依赖区域约束采用向量的∞范数约束的形式，从而使子问题变二次规划，同时使算法变得更实用。在通常假设条件下，证明了算法的整体收敛性和超线性收敛性。     

仿射变换内点信赖域方法求解变分不等式问题

本文研究了求解非线性约束变分不等式问题(VIP)的一个新的算法.利用KKT条件的非光滑方程形式,得到了与VIP等价的简单约束优化问题.提出了求解VIP的一类结合回代线搜索技巧的仿射变换内点信赖域算法.在较弱的条件下证明了算法具有整体收敛性,进一步在某些正则条件下,证明了算法具有超线性收敛速度.     

解P_0非线性互补问题的光滑Levenberg-Marquardt方法

本文研究了一个P0非线性互补问题.利用信赖域技术获得了求解该问题的光滑Levenberg-Marquardt算法,该算法在一定条件下具有全局性.利用局部误差界还获得了该算法的超线性和二次收敛.数值结果表明该算法是有效的.     

解不等式约束优化的新的序列线性方程组方法

提出一种新的序列线性方程组(SSLE)算法解非线性不等式约束优化问题.在算法的每步迭代,子问题只需解四个简化的有相同的系数矩阵的线性方程组.证明算法是可行的,并且不需假定聚点的孤立性、严格互补条件和积极约束函数的梯度的线性独立性得到算法的全局收敛性.在一定条件下,证明算法的超线性收敛率.     

无严格互补松驰条件的序列线性方程组新算法

该文通过构造特殊形式的有效集来逼近KKT点处的有效集，给出了一个任意初始点下的序列线性方程组新算法，并证明了该算法在没有严格互补松驰条件的情况下具有全局收敛性和一步超线性收敛性。      

约束最优化的一个使用非单调线搜索的超线性收敛算法

Abstract. In the paper, a new mixed algorithm combined with schemes of nonmonotone line search, the systems of linear equations for higher order modification and sequential quadratic programming for constrained optimizations is presented. Under some weaker assumptions,without strict complementary condition, the algorithm is globally and superlinearly convergent.     

Superlinear Convergence of a Newton-Type Algorithm for Monotone Equations

We consider the problem of finding solutions of systems of monotone equations. The Newton-type algorithm proposed in Ref. 1 has a very nice global convergence property in that the whole sequence of iterates generated by this algorithm converges to a solution, if it exists. Superlinear convergence of this algorithm is obtained under a standard nonsingularity assumption. The nonsingularity condition implies that the problem has a unique solution; thus, for a problem with more than one solution, such a nonsingularity condition cannot hold. In this paper, we show that the superlinear convergence of this algorithm still holds under a local error-bound assumption that is weaker than the standard nonsingularity condition. The local error-bound condition may hold even for problems with nonunique solutions. As an application, we obtain a Newton algorithm with very nice global and superlinear convergence for the minimum norm solution of linear programs.This research was supported by the Singapore-MIT Alliance and the Australian Research Council.     

An improved interior-type feasible QP-free algorithm for inequality constrained optimization problems

In this paper, an improved interior-type feasible QP-free algorithm for inequality constrained optimization problems is proposed. At each iteration, by solving three systems of linear equations with the same coefficient matrix, a search direction is generated. The algorithm is proved to be globally and superlinearly convergent under some mild conditions. Preliminary numerical results show that the proposed algorithm may be promising. Advantages of the algorithm include: the uniformly nonsingularity of the coefficient matrices without the strictly complementarity condition is obtained. Moreover, the global convergence is achieved even if the number of the stationary points is infinite.     

A MIXED SUPERLINEARLY CONVERGENT ALGORITHM WITH NONMONOTONE SEARCH FOR CONSTRAINED OPTIMIZATIONS

§ 1　 IntroductionConsider the optimization problemmin{ f(x) :gj(x)≤ 0 ,j∈ I,gj(x) =0 ,j∈ L,x∈ Rn} ,(1 )where f(x) ,gj(x) :Rn→R,j∈I∪L.I={ 1 ,2 ,...,m} ,L={ m 1 ,...,m p} .We know thatthe sequential quadratic programming(SQP) [1～ 4] is one of the mostef-ficient methods to solve problem(1 ) because of its superlinear convergence.In order toovercome the Maratos effect[5] ,SQP should solve two quadratic sub-programmings ateachiteration,which,however,causes the amountof computation…     

一类特征值反问题(IEP)的基于矩阵方程的Ulm型算法

应用求解算子方程的Ulm方法构造了求解一类矩阵特征值反问题（IEP）的新算法.所给算法避免了文献[Aishima K.，A quadratically convergent algorithm based on matrix equations for inverse eigenvalue problems，Linear Algebra and its Applications，2018，542：310-33]中算法在每次迭代中要求解一个线性方程组的不足，证明了在给定谱数据互不相同的条件下所给算法具有根收敛意义下的二次收敛性.数值实验表明本文所给算法在矩阵阶数较大时计算效果优于上文所给算法.     

A NEW CONSTRAINTS IDENTIFICATION TECHNIQUE-BASED QP-FREE ALGORITHM FOR THE SOLUTION OF INEQUALITY CONSTRAINED MINIMIZATION PROBLEMS

In this paper, we propose a feasible QP-free method for solving nonlinear inequality constrained optimization problems. A new working set is proposed to estimate the active set. Specially, to determine the working set, the new method makes use of the multiplier information from the previous iteration, eliminating the need to compute a multiplier function. At each iteration, two or three reduced symmetric systems of linear equations with a common coefficient matrix involving only constraints in the working set are solved, and when the iterate is sufficiently close to a KKT point, only two of them are involved. Moreover, the new algorithm is proved to be globally convergent to a KKT point under mild conditions. Without assuming the strict complementarity, the convergence rate is superlinear under a condition weaker than the strong second-order sufficiency condition. Numerical experiments illustrate the efficiency of the algorithm.     

一类LC1规划问题的超线性收敛ODE信赖域算法

In this paper, a new trust region algorithm for unconstrained LC1 optimization problems is given. Compare with those existing trust regiion methods, this algorithm has a different feature: it obtains a stepsize at each iteration not by soloving a quadratic subproblem with a trust region bound, but by solving a system of linear equations. Thus it reduces computational complexity and improves computation efficiency. It is proven that this algorithm is globally convergent and locally superlinear under some conditions.