摘要
通过将非线性LC1约束优化问题的KKT条件转化成半光滑方程组,提出了求解LC1约束优化问题的非精确广义牛顿算法.并给出了保证该算法超线性收敛的构造方法,使得算法得以实现.
A construct algorithm was proposed, which can confirm globally and superlinear convergence of the inexact generalized Newtion's method for nonlinear LC^1 constrained optimization problem, and the problem was solved through solving semismooth equations reformulated from KKT conditions.
出处
《苏州大学学报(自然科学版)》
CAS
2005年第4期92-94,共3页
Journal of Soochow University(Natural Science Edition)
关键词
LC^1约束优化问题
半光滑方程
非精确广义牛顿法
全局收敛
超线性收敛
LC^1 constrained optimization problem
semismooth equations
inexact generalized Newton's method
globally convergent
superlinear convergent