首页 | 本学科首页   官方微博 | 高级检索  
     

一种无恢复过程的SQP-滤子法
引用本文:左双勇 王祥玲 朱志斌. 一种无恢复过程的SQP-滤子法[J]. 应用数学, 2019, 32(1): 106-112
作者姓名:左双勇 王祥玲 朱志斌
作者单位:宜春幼儿师范高等专科学校初等教育学院;桂林电子科技大学数学与计算科学学院
基金项目:国家自然科学基金(11361018);广西自然科学基金资助项目(2014GXNSFFA118001);宜春市社科研究"十三五"规划项目(YCSK2018-115;YCSK2018-106)
摘    要:本文研究非线性不等式约束优化问题,构造一个新的SQP-滤子法.该方法将滤子技术有机融合到简金宝提出的可行SQP方法中,利用转轴运算的思想,产生一个近似积极约束集,当QP子问题不相容时,利用广义投影技术获得可行搜索方向.该算法既能避免罚函数的选择,又能避免常规滤子算法中的恢复算法,一定程度上简化了计算.最后,在合理的条件下,证明了算法的全局收敛性.

关 键 词:非线性不等式约束  转轴运算  广义投影技术  滤子技术
收稿时间:2018-03-06

An SQP-Filter Method Without Restoration Process
ZUO Shuangyong,WANG Xiangling,ZHU Zhibin. An SQP-Filter Method Without Restoration Process[J]. Mathematica Applicata, 2019, 32(1): 106-112
Authors:ZUO Shuangyong  WANG Xiangling  ZHU Zhibin
Affiliation:(Primary Education College,Yichun Early Childhood Teachers College,Yichun 330814,China;Department of Mathematics and Computational Science,Guilin University ofElectronic Technology,Guilin 541004,China)
Abstract:In this paper, we consider the nonlinear inequality constrained optimization problem.A new SQP-filter method is presented. In the algorithm, the filter technique is combined to the feasible SQP method which is proposed by Jinbao Jian. An approximate active constraint set is produced by the idea of pivoting operation. When the QP subproblem is incompatible, the feasible direction of search is obtained by generalized gradient projection method. So this method is effective to avoid the restoration algorithm in general filter algorithm and the difficulties in choosing penalty parameter. Therefore, the computational cost is reduced. The theoretical analysis shows that the algorithm is global convergent under some suitable conditions.
Keywords:Nonlinear inequality constrained optimization  Pivoting operation  Generalized projection technique  Filter technique
本文献已被 CNKI 维普 等数据库收录!
点击此处可从《应用数学》浏览原始摘要信息
点击此处可从《应用数学》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号