摘要
利用互补问题的Lagrange函数,给出了互补约束优化问题(MPCC)的一种新松弛问题.在较弱的条件下,新松弛问题满足线性独立约束规范.在此基础上,提出了求解互补约束优化问题的乘子松弛法.在MPCC-LICQ条件下,松弛问题稳定点的任何聚点都是MPCC的M-稳定点.无需二阶必要条件,只在ULSC条件下,就可保证聚点是MPCC的B-稳定点.另外,给出了算法收敛于B-稳定点的新条件.
By using the Lagrange function of the complementarity problem, a new relaxed problem of MPCC is given. We show that the linear independence constraint qualification holds for the new relaxed problem under some mild conditions. Based on this, a multiplier relaxed method for solving MPCC is presented. The limited point of stationary points of the relaxed problems is M-stationary point under the MPCC-LICQ. Without requiring the second-order necessary condition, the limited point is B-stationary point if the ULSC holds. At last, we propose a new condition for convergence to B- stationary point.
出处
《运筹学学报》
CSCD
北大核心
2014年第4期119-130,共12页
Operations Research Transactions
基金
国家自然科学基金(No.11261033)
内蒙古大学高层次人才引进科研启动基金(No.210143)
关键词
互补约束优化问题
LAGRANGE函数
上水平严格互补
B-稳定点
mathematical programs with complementarity constraints, Lagrangefunction, upper lever strict complementarity, B-stationary point