A hybrid conjugate gradient method with descent property for unconstrained optimization
A hybrid conjugate gradient method with descent property for unconstrained optimization
复制标题
DOI：
10.1016/j.apm.2014.08.008
复制
发表时间：
2015-02-01
影响因子：
5
通讯作者：
Jiang, Xianzhen
中科院分区：
文献类型：
--
作者：
Jian, Jinbao;Han, Lin;Jiang, Xianzhen
本文在前人共轭梯度法的基础上，提出了一种新的求解无约束优化问题的混合共轭梯度法。该方法在每次迭代中都能产生良好的方向，而且这种性质与步长线搜索无关。在Wolfe线搜索下，该方法具有全局收敛性.中等规模的数值实验和它们的性能曲线的报告，这表明该方法是有前途的。(C)2014爱思唯尔公司All rights reserved.
In this paper, based on some famous previous conjugate gradient methods, a new hybrid conjugate gradient method was presented for unconstrained optimization. The proposed method can generate decent directions at every iteration, moreover, this property is independent of the steplength line search. Under the Wolfe line search, the proposed method possesses global convergence. Medium-scale numerical experiments and their performance profiles are reported, which show that the proposed method is promising. (C) 2014 Elsevier Inc. All rights reserved.