摘要
给出一类求解非线性无约束优化问题的共轭梯度新算法。在强Wolfe-Powell线搜索下所给公式具有充分下降性,所给该新算法具有全局收敛性。
A new kind of conjugate gradient algorithm for solving nonlinear unconstrained optimization was proposed. This conjugate gradient method with strong Wolfe-Powell line search satisfies the sufficient descent property, and this new kind of conjugate gradient algorithm with strong Wolfe line search possess the global convergence property.
出处
《山东大学学报(理学版)》
CAS
CSCD
北大核心
2008年第5期54-57,62,共5页
Journal of Shandong University(Natural Science)
关键词
共轭梯度法
线搜索
充分下降性
全局收敛性
conjugate gradient method
line search
sufficient descent
global convergence