非单调信赖域方法求解无约束非光滑优化问题

提出了非单调信赖域算法求解无约束非光滑优化问题,并和经典的信赖域方法作比较分析。同时,设定了一些条件,在这些假设条件下证明了该算法是整体收敛的。数值实验结果表明,非单调策略对无约束非光滑优化问题的求解是行之有效的,拓展了非单调信赖域算法的应用领域。     

一类新的带记忆模型的非单调信赖域算法

本文就无约束优化问题提出了一个带记忆模型的非单调信赖域算法。与传统的非单调信赖域算法不同,文中的信赖域子问题的逼近模型为记忆模型,该模型使我们可以从更全面的角度来求得信赖域试探步,从而避免了传统非单调信赖域方法中试探步的求取完全依赖于当前点的信息而过于局部化的困难。文中提出了一个带记忆模型的非单调信赖域
域算法,并证明了其全局收敛性。     

求解半定规划的新算法

为了提高求解半定规划问题的运算效率,提出了一种新的求解半定规划的非单调信赖域算法。将半定规划的最优性条件转化为无约束优化问题,并构造无约束优化问题的信赖域子问题,修正信赖域半径的校正条件,当初始搜索点处于峡谷附近时仍能搜索到全局最优解。实验结果表明,对于小规模和中等规模的半定规划问题,该算法的迭代次数都比经典的内点算法少,运行速度快。     

求解半定规划的新算法

为了提高求解半定规划问题的运算效率,提出了一种新的求解半定规划的非单调信赖域算法。将半定规划的最优性条件转化为无约束优化问题,并构造无约束优化问题的信赖域子问题,修正信赖域半径的校正条件,当初始搜索点处于峡谷附近时仍能搜索到全局最优解。实验结果表明,对于小规模和中等规模的半定规划问题,该算法的迭代次数都比经典的内点算法少,运行速度快。     

求解二次规划逆问题的非单调信赖域算法

为了提高求解二次规划逆问题的速度,提出了针对求解该问题的非单调信赖域算法.为了降低问题的复杂度,将二次规划逆问题转换为决策变量相对较少的对偶问题,采用增广Lagrange法构造对偶问题的子问题,并通过引入光滑函数将子问题转换为无约束优化问题,利用非单调信赖域算法进行求解.数值实验结果表明,该算法的迭代次数比牛顿算法、Gauss回代交替方向法少,运行速度快.因此,对于大规模二次规划逆问题,该算法更加有效.     

一种求解半定规划的非单调信赖域算法

提出一种求解半定规划的非单调信赖域算法。利用推广至矩阵域的光滑Fischer-Burmeister函数,转化半定规划的最优性条件,改写半定规划的中心路径,得到与其等价的无约束优化问题的非线性可微光滑方程组,在求解信赖域子问题时,利用当前迭代点的一阶梯度信息,给出信赖域半径的选取机制。仿真结果表明,与经典的内点算法相比,对于一般规模(n, m≤30)的半定规划问题,该算法的运行速度较快。对于大规模的半定规划问题(n, m＞30),该算法更适合处理Norm min、Lovasz这2类问题。     

基于信赖域Newton 算法的ELM网络

针对极端学习机(ELM)网络伪逆输出权值计算方法的运算复杂度制约其训练速度问题,提出一种基于信赖域Newton算法的新型ELM网络(TRON-ELM),并采用信赖域Newton算法求解ELM网络的输出权值.该算法首先构造一个ELM网络代价函数的Newton方程,并将其作为一个无约束优化问题,采用共轭梯度法求解,避免了求代价函数Hessian矩阵逆的运算,提高了训练速度,信赖域条件的存在保证了算法的整体收敛性.仿真实验结果验证了所提出方法的有效性.     

一种非线性互补问题的信赖域算法收敛性

针对非线性互补问题求解困难,利用信赖域算法,并结合极大熵函数法给出该类问题的一种新的有效算法．该算法首先利用极大熵函数将非线性互补问题转化为一个无约束最优化问题,然后应用信赖域算法来优化该问题,并在一定条件下证明该算法具有全局收敛性。数值算例表明算法的有效性。     

求解互补支持向量机的非单调信赖域算法

求解支持向量机的核心问题是对一个大规模凸二次规划问题进行求解。基于支持向量机的修正模型,得到一个与之等价的互补问题,利用Fischer-Burmeister互补函数,从一个新的角度提出了求解互补支持向量机的非单调信赖域算法。新算法避免了求解Hesse矩阵或矩阵求逆运算,减少了工作量,提高了运算效率。在不需要任何假设的情况下,证明算法具有全局收敛性。数值实验结果表明,对于大规模非线性分类问题,该算法的运行速度比LSVM算法和下降法快,为求解SVM优化问题提供了一种新的可行方法。     

信赖域共轭梯度法求解二次规划逆问题

为了有效地求解二次规划逆问题,提出了一种求解其对偶问题的子问题的光滑化信赖域共轭梯度法。该方法采用增广拉格朗日法求解其对偶问题,引入光滑函数将对偶问题的子问题转换成连续的无约束优化问题,将信赖域法与共轭梯度法结合,设计出求解二次规划逆问题的算法流程。数值实验结果表明,该方法可行且有效,与牛顿法相比,更适合求解大规模问题。     

用改进的信赖域方法求解二次插值模型

一种改进的信赖域方法被用来解无约束最优化问题,当目标函数的导数信息不可利用或者求解目标函数的导数代价太大。通常,考虑用二次插值模型来逼近目标函数,并且用传统的信赖域方法求解这个二次模型。传统的信赖域方法将被改进,并且形成两个改进的信赖域子问题。改进的信赖域方法的创新点在于:求解二次模型在一个参数化的信赖域中,修改这个模型在另一个参数化的信赖域当中。在这两个新的信赖域中,可以分别很快地找到一个好的下降方向和一个具有均衡性的插值点。这个改进的方法不但节省了函数值计算次数而且提高了解的精度。实验结果表明,针对测试问题,提出的方法的确是优于传统的信赖域方法的。     

Trust region methods for solving the optimal output feedback design problem

We consider the problem of designing feedback control laws when a complete set of state variables is not available. For linear autonomous control systems with quadratic performance criterion, the design problem consists of choosing an appropriate static output feedback (SOF) gain matrix according to a certain objective function. The corresponding non-linear matrix optimization problem can be interpreted as an equality constrained minimization problem. For solving this problem, we propose a constrained trust region (CTR) algorithm, which presents a new and efficient numerical approach for this problem class. On the other hand, based on the formulation used in the past, the SOF problem can be also interpreted as an unconstrained programming problem. Thus, based on this interpretation, we also develop an unconstrained trust region (UTR) method. Finally, several numerical examples for optimal SOF problems demonstrate the applicability of the considered algorithm.     

An adaptive nonmonotone trust region algorithm

Based on an eigenvalue analysis conducted on the scaled memoryless quasi-Newton updating formulas BFGS and DFP, an adaptive choice for the trust region radius is proposed. Then, using a trust region ratio obtained from a nonmonotone line search strategy, an adaptive nonmonotone trust region algorithm is developed. Under proper conditions, it is briefly shown that the proposed algorithm is globally and locally superlinearly convergent. Numerical experiments are done on a set of unconstrained optimization test problems of the CUTEr collection, using the Dolan–Moré performance profile. They show efficiency of the proposed algorithm.     

A variational method for unconstrained optimization problems

We introduce a new method for solving an unconstrained optimization problem. The method is based on the solution of a variational inequality problem and may be considered a modified trust region algorithm. We prove the convergence of the method and present some numerical results.     

Optimization of unconstrained problems using a developed algorithm of spectral conjugate gradient method calculation

This paper presents a numerical investigation of the spectral conjugate directions formulation for optimizing unconstrained problems. A novel modified algorithm is proposed based on the conjugate gradient coefficient method. The algorithm employs the Wolfe inexact line search conditions to determine the optimum step length at each iteration and selects the appropriate conjugate gradient coefficient accordingly. The algorithm is evaluated through several numerical experiments using various unconstrained functions. The results indicate that the algorithm is highly stable, regardless of the starting point, and has better convergence rates and efficiency compared to classical methods in certain cases. Overall, this research provides a promising approach to solving unconstrained optimization problems.     

A line up evolutionary algorithm for solving nonlinear constrained optimization problems

In this work a complete framework is presented for solving nonlinear constrained optimization problems, based on the line-up differential evolution (LUDE) algorithm which is proposed for solving unconstrained problems. Linear and/or nonlinear constraints are handled by embodying them in an augmented Lagrangian function, where the penalty parameters and multipliers are adapted as the execution of the algorithm proceeds. The LUDE algorithm maintains a population of solutions, which is continuously improved as it thrives from generation to generation. In each generation the solutions are lined up according to the corresponding objective function values. The position's in the line are very important, since they determine to what extent the crossover and the mutation operators are applied to each particular solution. The efficiency of the proposed methodology is illustrated by solving numerous unconstrained and constrained optimization problems and comparing it with other optimization techniques that can be found in the literature.