摘要
本文推广了无约束最优化方法中采用曲线搜索这一结果,从而给出了一类非线性约束条件下来用曲线搜索的可行方向法。并且在一定的条件下,我们证明了此算法类是全局收敛的. 关键词:非线性约束;;下降可行方向对;;曲线搜索;;可行方向法;;收敛性。
In this paper, we extend the curvilinear search algorithms ofunconstrainted optimization. Thus,we give a class of feasible directionalgorithm for nonlinear programming with nonlinear constraints by usingcurvilinear search.Under some weaker conditions,we prove that this class ofalgorithms is globally convergent.
出处
《工程数学学报》
CSCD
1989年第3期63-69,共7页
Chinese Journal of Engineering Mathematics
关键词
非线性约束
曲线搜索
可行方向法
Nonlinear constraint
Direct pair of feasible descent
Curvi linear search
Feasible direction
method
Convergence
