| 摘要: |
| 通过适当修正Fletcher Reeves(FR)方法，提出了一类修正FR共轭梯度法方法(MFR*),并证明了MFR*方法在强Wolfe线搜索下具有充分下降条件和全局收敛性. |
| 关键词: 无约束优化 共轭梯度法 充分下降性 全局收敛性 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| A Class of Modified FR Method with Strong Wolfe Line Search |
|
ZHANG Peng
|
| Abstract: |
| This paper proposes a class of modified FR method (MFR*) by modifying Fletcher Reeves (FR), and proves MFR* with strong Wolfe line search satisfies sufficient descent condition and global convergence. |
| Key words: unconstrained optimization conjugate gradient method sufficient descent condition global convergence |
