| 摘要: |
| 非线性共轭梯度法是解决大规模优化问题的一种非常有效的方法，提出了一个修正的PRP方法，该参数带有干扰因子，并证明了这一新的参数具有非负性，且在适当条件下，采用Wolfe线搜索，证明该算法具有全局收敛性. |
| 关键词: 共轭梯度法 强Wolfe条件 Wolfe条件 干扰因子 全局收敛性 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| Global Convergence Property of a Modified PRP Conjugate Gradient Method with Disturbance Factor |
|
CHEN Hong min
|
| Abstract: |
| The nonlinear conjugate gradient method is a very effective iterative method for solving large scale optimal problems.In this paper，a modified PRP conjugate gradient method with disturbance factor is proposed and non negative property of the formula is proved.Under suitable conditions，the global convergence of the algorithm with the Wolfe line search is discussed. |
| Key words: conjugate gradient method strong Wolfe condition Wolfe condition disturbance factor global convergence |
