几个修正的DY共轭梯度法  

Several Modified DY Conjugate Gradient Methods

作　　者：陈恩 CHEN En(School of Mathematics Science,Chongqing Normal University,Chongqing 401331,China)

机构地区：[1]重庆师范大学数学科学学院,重庆401331

出　　处：《湖北民族学院学报（自然科学版）》2018年第2期169-171,共3页Journal of Hubei Minzu University(Natural Science Edition)

摘　　要：共轭梯度法是求解无约束优化问题的一个非常有用的方法,特别是解决大规模问题,经典的DY法在Wolfe线搜索下就能保证下降性和全局收敛性.本文提出了四个修正的DY法,其中两种方法在Wolfe线搜索下具有下降性与全局收敛性;另两种方法不依赖线搜索具有充分下降性,在Wolfe线搜索下对一般函数具有全局收敛性.Conjugate gradient method is a very useful method for solving unconstrained optimization problems,especially for solving large-scale problems.The classical DY method can guarantee the descent and global convergence under the Wolfe line search.This paper presents four modified DY methods：Two search methods have the descent property and global convergence in the Wolfe line;the other two methods satisfy the sufficient descent without depending on the line search and have global convergence under the Wolfe line search.

关 键 词：无约束优化问题 共轭梯度法 WOLFE线搜索 全局收敛性 

分 类 号：O224[理学—运筹学与控制论]