摘要
本文给出了一类关于非线性约束条件下的可行方向法。在较简单的假设之下,我们证明了算法具有全局收敛性。特别在本文中,我们利用此类算法和两个已有的线性约束下的梯度投影法导出了两个较简的非线性约束条件下收敛的梯度投影法。
A Class of feasible direction methods for nonlinear programming with nonlinear con-straints is given, which is proved to be globlly convergent under simple conditions. Using this class and two available gradient projection methods for nonlinear programming with linear constraints, two simpler gradient projection methods for nonlinear program-ming with nonlinear constraints are derived.
出处
《高校应用数学学报(A辑)》
CSCD
北大核心
1990年第4期457-467,共11页
Applied Mathematics A Journal of Chinese Universities(Ser.A)