凸约束优化问题的杂交三阶投影HS-PRP方法  

A Hybrid Three-Term Projected HS-PRP Method for Optimization with Convex Constraint

作　　者：周姣利 

机构地区：[1]长沙理工大学数学与统计学院,湖南 长沙

出　　处：《应用数学进展》2022年第8期5750-5759,共10页Advances in Applied Mathematics

摘　　要：本文提出了一种杂交三阶投影HS-PRP共轭梯度法求解凸约束优化问题并证明了该算法的全局收敛性,该方法是求解无约束优化问题的三阶HS共轭梯度法的推广。数值实验结果表明,该算法是有效的。In this paper, we propose a hybrid third-term projected HS-PRP conjugate gradient method for solving convex constrained optimization problems and establish its global convergence, which is a generalization of the third-term HS conjugate gradient method for unconstrained optimization. Numerical experimental results show that the algorithm is effective.

关 键 词：投影 共轭梯度法 线搜索 全局收敛 

分 类 号：O17[理学—数学]