计算机工程Issue(9)：233-236,4.DOI:10.3969/j.issn.1000-3428.2013.09.052
一种求解半定规划的非单调信赖域算法
A Nonmonotonic Trust Region Algorithm for Solving Semidefinite Programming
摘要
Abstract
A nonmonotonic trust region algorithm for solving Semidefinite Programming(SDP) is proposed in this paper. The equivalent smoothing equations of the optimal condition are obtained by exploiting the Fischer-Burmeister function that is extended to the matrix domain, and the center of the path of SDP is rewritten. The algorithm makes full use of first-order gradient information of the current iteration point to solve the trust region subproblem, and a new trust region radius selection mechanism is proposed. Simulation results show that the algorithm runs faster than the classical interior point algorithm for general scale semidefinite programming problems(n, m≤30), for large-scale semidefinite programming problem(n, m>30) the algorithm is suitable for handling Norm min, Lovasz these two kinds of problems.关键词
半定规划/信赖域算法/非单调策略/内点算法/Fischer-Burmeister函数/无约束优化问题Key words
Semidefinite Programming(SDP)/trust region algorithm/nonmonotonic strategy/interior algorithm/Fischer-Burmeister function/unconstrained optimization problem分类
信息技术与安全科学引用本文复制引用
高雷阜,于冬梅,张兴涛..一种求解半定规划的非单调信赖域算法[J].计算机工程,2013,(9):233-236,4.基金项目
辽宁省教育厅青年基金资助项目(L2012105)；教育部高等学校博士学科点专项科研基金资助项目(20102121110002) （L2012105）
