机构地区: 湖南大学数学与计量经济学院
出 处: 《应用数学》 2008年第S1期6-11,共6页
摘 要: 本文在求解无约束最优化问题的MFR共轭梯度法和MPRP共轭梯度法中引入两种非单调线性搜索技术.我们证明在适当条件下采用非单调线性搜索的MFR算法和MPRP算法具有全局收敛性.数值结果表明非单调线性搜索具有优越性. In this paper,we introduce two kinds of nonmonotone line searches to the MFR and the MPRP conjugate gradient method for solving unconstrained optimization problems.Under mild conditions,we prove that the MFR algorithm and MPRP algorithm with nonmonotone line searches are globally convergent.We also report some numerical results.The results show the methods with nonmonotone line searches perform well.