A new double trust regions SQP method without a penalty function or a filter

IF 2.5 3区数学 Q1 MATHEMATICS, APPLIED

Computational & Applied Mathematics Pub Date : 2012-08-28 DOI:10.1590/S1807-03022012000200011

Xiaojing Zhu, D. Pu

引用次数: 2

Abstract

A new trust-region SQP method for equality constrained optimization is considered. This method avoids using a penalty function or a filter, and yet can be globally convergent to first-order critical points under some reasonable assumptions. Each SQP step is composed of a normal step and a tangential step for which different trust regions are applied in the spirit of Gould and Toint [Math. Program., 122 (2010), pp. 155-196]. Numerical results demonstrate that this new approach is potentially useful. Mathematical subject classification: 65K05, 90C30, 90C55.

查看原文本刊更多论文

一种新的不带惩罚函数和过滤器的双信任域SQP方法

提出了求解等式约束优化问题的一种新的信任域SQP方法。该方法避免了使用惩罚函数或滤波器，并且在合理的假设下可以全局收敛到一阶临界点。每个SQP步骤由一个正规步骤和一个切向步骤组成，根据Gould和Toint [Math]的精神，对这些步骤应用不同的信任区域。程序。， 122 (2010)， pp. 155-196]。数值结果表明，该方法具有潜在的实用价值。数学学科分类:65K05、90C30、90C55。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Computational & Applied Mathematics Mathematics-Computational Mathematics

CiteScore

4.50

自引率

11.50%

发文量

352

审稿时长

>12 weeks

期刊介绍： Computational & Applied Mathematics began to be published in 1981. This journal was conceived as the main scientific publication of SBMAC (Brazilian Society of Computational and Applied Mathematics). The objective of the journal is the publication of original research in Applied and Computational Mathematics, with interfaces in Physics, Engineering, Chemistry, Biology, Operations Research, Statistics, Social Sciences and Economy. The journal has the usual quality standards of scientific international journals and we aim high level of contributions in terms of originality, depth and relevance.