摘要
本文讨论带有线性非负约束的凸不可微规划问题,给出了解这类规划的既约次梯度算法.所给出的算法采用不精确的线性搜索,而且证明了算法的收敛性.
In this paper, we discuss the non—differentiable optmization problem with linear constraints and nonnegative variables and give a subgradient algorithm for this problem. The algorithim presented is a descent and iterative method by nonexact line search. We have proved the global convergence of the algorithm.
出处
《山东师范大学学报（自然科学版）》
CAS
1993年第1期4-9,共6页
Journal of Shandong Normal University(Natural Science)
关键词
不可微规划
梯度算法
既约次梯度法
Linear nonnegative constraints
convex nonsmooth programming
subgradient algorithm
