Armijo型线搜索下一种共轭梯度法的收敛性

对无约束非线性规划问题,本文分别在两种不同的Armijo型线搜索下证明了Liu-Storey共轭梯度法的所有搜索方向都是充分下降的,并进一步证明了该算法是全局强收敛的。对另一种放松了函数值下降条件可以获得更大步长的Armijo型线搜索,本文还证明了该算法是全局强收敛的。     

结合广义Armijo步长搜索的一类新的共轭度算法及其收敛特征

对求解无约束规划的共轭梯度算法中共轭梯度方向中的参数给了一个假设条件。从而确定它的一个取值范围，使其在此范围内取值均能保证共轭梯度方向是目标函数的充分下降方向。提出了一类新的共轭梯度算法，在去掉迭代点列有界和广义Armijo步长搜索下讨论了算法的全局收敛性。同时给出了具有好的收敛性质和较快收敛速度的FR，PR，HS共轭梯度法的修正形式。数值例子表明新算法比Armijo搜索下的FR，PR，HS共轭梯算法更稳定更有效。算法需要较小的存储。特别适于求解大规模无约束最优化问题。     

一个充分下降的修正 PRP 型谱共轭梯度法

谱共轭梯度法是共轭梯度法的一种重要延拓,可以通过共轭参数和谱参数二维度调整,使得所设计算法的搜索方向满足某一预设条件,比如充分下降条件或共轭条件等。谱参数和共轭参数的设计是谱共轭梯度法的两大核心工作,决定方法的收敛性和数值效果。基于 PRP 方法,构造了一个修正的 PRP 型共轭参数,该共轭参数不仅保持了 PRP 公式的结构和性能,而且具有 FR 方法的收敛性质。利用充分下降条件取定一个谱参数,与修正的 PRP 型共轭参数结合,建立一个新的谱共轭梯度算法。该算法不依赖于任何线搜索就可以满足充分下降条件。常规假设条件下,采用强 Wolfe 线搜索准则产生步长,证明了新算法的全局敛性。通过 100 个算例对该算法进行数值测试并与其他五个算法进行比较,同时采用性能图对数值结果进行直观展示,结果表明该算法是有效的。     

结合广义Armijo步长搜索的一类记忆梯度算法及其收敛特征

本文对求解无约束规划的超记忆梯度算法中线搜索方向中的参数,给了一个假设条件,从而确定了它的一个新的取值范围,保证了搜索方向是目标函数的充分下降方向,由此提出了一类新的记忆梯度算法.并在去掉迭代点列有界和广义Armijo步长搜索下,讨论了算法的全局收敛性,且给出了结合形如共轭梯度法FR,PR,HS的记忆梯度法的修正形式,数值实验表明,新算法比Armijo线搜索下的FR,PR,HS共轭梯度法和超记忆梯度法更稳定、更有效.     

一类Armijo搜索下的混合HS-PRP共轭梯度法?

为有效求解大规模无约束优化问题,本文基于HS方法和PRP方法,提出了一类新的混合共轭梯度法。该方法在每步迭代中都不依赖于函数的凸性和搜索条件而自行产生充分下降方向。在精确搜索下,本文算法将还原为标准的PRP方法。在适当的条件下,获证了该法在Armijo搜索下,即使求解非凸函数极小化的问题,算法也具有全局收敛性。同时,数值实验表明本文算法可以有效求解优化测试问题。     

一个新的谱共轭梯度法

谱共轭梯度法含有两个方向调控参数,是求解无约束优化问题的一类有效方法.本文给出一对参数公式以构建新的谱共轭梯度法,该方法在精确线搜索下与标准FR方法等价,在Wolfe线搜索下具有类似标准DY方法的内在性质.我们证明了采用Wolfe线搜索的新算法在每一次迭代中均产生下降方向,并且具有全局收敛性.数值实验结果表明,新算法数值稳定、有效,适合于求解大规模无约束优化问题.     

包含共轭下降法的一类无约束优化方法的全局收敛性

共轭下降法是由Fletcher(1987)提出的一个共轭梯度法，提出了包含共轭下降法的一类无约束优化方法，并采用Dai Yuhong和Yuan Yaxinag(1996)建立的一个非精确经搜索模型，给出了保证这类方法具有下降性和全局收敛性的搜索条件，文中得到的有关共轭下降法的收敛条件与Dai Yuhong和Yuan Yaxiang(1996)给出的收敛条件一致。     

Wolfe线搜索下一个新的全局收敛共轭梯度法

共轭梯度法是求解大规模无约束优化问题的有效方法之一,其研究十分活跃.本文给出了一个新的共轭梯度法公式,新公式在精确线搜索下与DY公式等价.基于新公式,采用Wolfe非精确线搜索确定步长,本文设计了一个新的共轭梯度算法,并证明了新算法的下降性和全局收敛性.数值试验结果表明所设计新算法是有效的.     

一族共轭梯度法的全局收敛性

提出了求解无约束优化问题的一族共轭梯度法，这族方法包古Fkcher提出的共轭下降法．文中证明了一种非精确线性搜索条件能够保证这族方法的下降性和全局收敛性，其收敛结果与Dai和Yuan 1996年给出的关于共轭下降法的相一致．     

线搜索下带误差项的Dai-Yuan共轭梯度算法

本文在共轭梯度不能精确计算的情况下,采用Wolfe或Armijo步长规则研究了带误差项的Dai-Yuan(abbr．DY)共轭梯度法,我们的方法的一个很重要的特征就是步长不一定趋于零。这种特征使得我们的分析对许多实际问题很有用。我们在很一般的假设条件下证明了算法的全局收敛性。最后给出了数值算例。     

共轭梯度法的全局收敛性

探讨了在强Wolfe搜索规则下，与βk^PR相关的算法的收敛性，在不需要假设目标函数为凸的情况下，证明了充分下降及算法的全局收敛性。     

A gradient‐only line search method for the conjugate gradient method applied to constrained optimization problems with severe noise in the objective function

A new implementation of the conjugate gradient method is presented that economically overcomes the problem of severe numerical noise superimposed on an otherwise smooth underlying objective function of a constrained optimization problem. This is done by the use of a novel gradient‐only line search technique, which requires only two gradient vector evaluations per search direction and no explicit function evaluations. The use of this line search technique is not restricted to the conjugate gradient method but may be applied to any line search descent method. This method, in which the gradients may be computed by central finite differences with relatively large perturbations, allows for the effective smoothing out of any numerical noise present in the objective function. This new implementation of the conjugate gradient method, referred to as the ETOPC algorithm, is tested using a large number of well‐known test problems. For initial tests with no noise introduced in the objective functions, and with high accuracy requirements set, it is found that the proposed new conjugate gradient implementation is as robust and reliable as traditional first‐order penalty function methods. With the introduction of severe relative random noise in the objective function, the results are surprisingly good, with accuracies obtained that are more than sufficient compared to that required for engineering design optimization problems with similar noise. Copyright © 2004 John Wiley & Sons, Ltd.     

An arc-length method including line searches and accelerations

This paper describes a method for introducing line searches into the arc-length solution procedure. Such line searches may be used at each iteration to calculate an optimum scalar step-length which scales the normal iterative vector. In practice, a loose tolerance is provided so that on many iterations the line searches are avoided. However on ‘difficult iterations’, the line searches are shown to lead to a substantial improvement in the convergence characteristics. A simple single-parameter acceleration is also developed using line search concepts. The new arc-length method is applied to both the geometrically nonlinear analysis of shallow shells and the materially nonlinear analysis of reinforced concrete beams and slabs. Significant improvements are demonstrated in relation to the standard arc-length method.     

An application of the arc length method involving concrete cracking

To overcome numerical difficulties in highly non-linear materials Crisfield⁶ proposed a solution procedure involving the use of a constant arc length solution scheme with line searches and accelerations. This paper uses the arc length method with line searches to increase the stability and substantially reduce the cpu time required for convergence in the finite element modelling of reinforced concrete. The arc length method is compared with the modified Newton–Raphson solution scheme and it was found that substantial savings in computer resources are obtained by using the arc length method. On difficult iterations the use of line searches is shown to lead to substantial improvement, with the best results occurring when using a slack tolerance of ψ = 0.8.     

求解非线性等式约束优化问题的共轭梯度投影算法

利用投影矩阵,对求解无约束规划的共轭梯度算法中的参数βk给一限制条件确定βk的取值范围,以保证得到目标函数的共轭梯度投影下降方向,建立了求解非线性等式约束优化问题的共轭梯度投影算法,并证明了算法的收敛性。数值例子表明算法是有效的。     

Line source and point source diffraction by a reactive step

The diffraction of a line and a point source by a reactive step joined by two half planes where each half plane and step are characterized by different surface reactances have been studied. The problem is solved by using the Wiener–Hopf technique and the Fourier transform. The scattered field in the far zone is determined by the method of steepest descent. Graphical results for the line source are also presented. It is observed that if the source is shifted to a large distance the results of the line source differ from those of Buyukaksoy and Birbir [Int. J. Eng. Sci. 1997, 35, 311–319] by a multiplicative factor to the part of the scattered field containing the effects of incident and reflected waves. Subsequently, the point source diffraction is examined using the results obtained for a line source diffraction.     

基于TRIZ进化树的滚筒洗衣机柔顺剂供给装置创新设计

利用进化树可以进行产品预测,并作出专利规避.进化树的画法是以物体的分割路线为主线,找出其中需要改进的阶段,进一步寻找可用的进化路线,最终形成进化树,从而在进化树中找出解决问题的方案.在立式滚筒洗衣机柔顺剂供给装置的改进创新中,利用进化树,根据系统组件裁剪进化路线和物体的几何进化路线的提示,找到了新的解决方案,对原来的四槽供给装置改进为单个三角棱柱供给装置,有效规避了现有的专利.