| 摘要: |
| 将一类Wolfe类线搜索模型的LS搜索模型与文献[10]提出的修改Broyden族(MBC1和M BC2)相结合,得到M BC1算法和MBC2算法,并证明MBC1算法和MBC2算法在LS搜索模型下具有全局收敛性. |
| 关键词: 无约束最优化 拟牛顿方程 Broyden族 全局收敛性 |
| DOI： |
| 投稿时间：2005-04-18 |
| 基金项目:国家自然科学基金(10161002);广西自然科学基金(0135004)联合资助 |
|
| Global Convergence of the Modified Broyden's Family with a Class of Inexact Line Searches |
|
WEI Zeng-xin, XIE Pin-jie
|
| (Department of Mathematics and Information Science, Guangxi University, Nanning, Guangxi, 530004, China) |
| Abstract: |
| Using the combinations of a class Wolfe-type inexact line search and a modified Broyden's family proposed by Reference[10], we derive a MBC1 algorithm and a MBC2 algorithm.Under suitable conditions, we prove that this two algorithms are global convergence with inexact linesearch. |
| Key words: unconstrained optimization quasi-Newton equation Broyden's family global convergence |
