带有混合约束的二次半定规划的内点算法  

An Interior-point Algorithm for Quadratic Semi-definite Programming with Mixed Constraints

作　　者：黎博[1] 田志远[1] 汪雪萍[1] 杨婷婷[1] 

机构地区：[1]青岛大学数学科学学院

出　　处：《青岛大学学报（自然科学版）》2011年第4期20-24,共5页Journal of Qingdao University(Natural Science Edition)

基　　金：山东省高等学校科技计划项目(JI0LA05)

摘　　要：研究带有混合约束的二次半定规划问题的内点算法。首先给出该问题的对偶问题和一种障碍函数,并建立相应的Lagrange函数,以此为基础给出内点算法,最后分析并证明了算法的全局收敛性。数值试验表明该算法是有效的。The interior-point algorithm for quadratic semi-definite programming with linear equality con- straints and inequality constraints is studied. First the dual problem and the barrier function are given and the corresponding Lagrange function is developed, then an interior-point algorithm for solving this pro- gramming is studied. Finally it＇s proved that the algorithm is executable and convergent. Numerical exper- iments show that the algorithm is effective.

关 键 词：半定规划 内点算法 收敛性 

分 类 号：O221.2[理学—运筹学与控制论]