一种解大规模无约束优化问题的BB型算法  

A BB algorithm for solving large-scale unconstrained optimization problems

作　　者：高健 欧宜贵[1] Gao Jian;Ou Yigui(School of Science,Hainan University,Haikou 570228,China)

机构地区：[1]海南大学理学院,海南海口570228

出　　处：《海南大学学报（自然科学版）》2023年第3期239-248,共10页Natural Science Journal of Hainan University

基　　金：国家自然科学基金(11961018);海南省普通高等学校研究生创新科研课题(Qhys2021-207)。

摘　　要：基于BB算法和IMPBOT算法的思想,结合修正割线方程得到2种新的BB步长,并利用Armijo线搜索,提出一种求解大规模无约束优化问题的BB型混合算法.在适当的条件下对其收敛性证明,并通过大量的数值试验来验证所提算法的有效性.In the report,based on the IMPBOT algorithm,the BB algorithm combined with the modified secant equation,two new BB step sizes were obtained,and Armijo line search was used to propose a BB-type hybrid method for solving large-scale unconstrained optimization problems.The convergence of the proposed algorithm was proved under appropriate conditions,a large number of numerical experiments were performed to verify the effectiveness of the proposed algorithm.

关 键 词：IMPBOT算法 BB算法 修正割线方程 大规模优化 数值实验 

分 类 号：O221.2[理学—运筹学与控制论]