摘要
针对具有等式约束的非线性最优化问题,提出了一类具有充分下降特性的投影Dai-Yuan共轭梯度法.在每次迭代过程中,算法均可得到充分下降的搜索方向.在适当条件下,证明了算法产生的搜索方向为可行下降方向,分析了算法的全局收敛性.数值结果表明算法是可行的、有效的.
In this paper,aprojected Dai-Yuan conjugate gradient method is proposed for nonlinear optimization problem with linear equality constraints.The direction of this paper is sufficient descent at each iteration.Under some suitable condition,the direction of the algorithm is feasible and descent.Furthermore,the global convergence is proved.Numerical results show that the method is feasible and effective.
作者
许春玲
孙颖异
李健
孙中波
XU Chun-ling;SUN Ying-yi;LI Jian;SUN Zhong-bo(Public Computer Department,College of Humanities and Sciences of Northeast Normal University,Changchun 130117,China;College of Information Technology,Jinlin Agriculture University,Changchun 130118,China;School of Electric and Electronic Engineering,Changchun University of Technology,Changchun 130012,China)
出处
《东北师大学报（自然科学版）》
CAS
北大核心
2019年第2期39-44,共6页
Journal of Northeast Normal University(Natural Science Edition)
基金
国家自然科学基金资助项目(61873304,51775054,11701209)
吉林省科技发展计划项目(20190302025GX,20170204067GX,20180201105GX)
吉林省智能机器人与视觉测控技术工程实验室建设项目(2019C010)
关键词
非线性等式约束优化
Dai-Yuan共轭梯度法
全局收敛性
充分下降方向
nonlinear optimization with linear equality constraints
Dai-Yuan conjugate gradient method
global convergence
sufficient descent direction
