An ε-generalized gradient projection method for nonlinear minimax problems

被引：7
|
作者：
Ma, Guo-Dong [1 ,2 ]
Jian, Jin-Bao [2 ]
机构：
[1] Shanghai Univ, Dept Math, Shanghai 200444, Peoples R China
[2] Yulin Normal Univ, Sch Math & Informat Sci, Yulin 537000, Guangxi, Peoples R China
来源：
NONLINEAR DYNAMICS | 2014年 / 75卷 / 04期
基金：
中国国家自然科学基金;
关键词：
Nonlinear minimax problems; epsilon-generalized gradient projection method; Global convergence; Strong convergence; SQP ALGORITHM; LINE SEARCH;
D O I：
10.1007/s11071-013-1095-1
中图分类号：
TH [机械、仪表工业];
学科分类号：
0802 ;
摘要：
In this paper, combining the techniques of epsilon-generalized gradient projection and Armjio's line search, we present a new algorithm for the nonlinear minimax problems. At each iteration, the improved search direction is generated by an epsilon-generalized gradient projection explicit formula. Under some mild assumptions, the algorithm possesses global and strong convergence. Finally, some preliminary numerical results show that the proposed algorithm performs efficiently.
引用
收藏
页码：693 / 700
页数：8
相关论文
共 50 条
12345