Journal of Computer Applications ›› 2014, Vol. 34 ›› Issue (1): 182-184.DOI: 10.11772/j.issn.1001-9081.2014.01.0182
• Network and distributed techno • Previous Articles Next Articles
YU Dongmei,GAO Leifu
Received:
Revised:
Online:
Published:
Contact:
于冬梅,高雷阜
通讯作者:
作者简介:
基金资助:
教育部高校博士学科点专项科研基金联合资助项目;国家自然科学基金资助项目;辽宁省教育厅基金资助项目
Abstract: In order to improve the operational efficiency of SemiDefinite Programming (SDP), a new nonmonotonic trust region algorithm was proposed. The SDP problem and its duality problem were transformed into unconstrained optimization problem and the trust region subproblem was constructed, the trust region radius correction condition was modified. When the initial search point was near the canyon, the global optimal solution still could be found. The experimental results show that the number of iterations of the algorithm is less than the classical interior point algorithm for small and medium scale semidefinite programming problems, and the proposed algorithm works faster.
Key words: SemiDefinite Programming (SDP), trust region algorithm, nonmonotonic strategy, interior algorithm, unconstrained optimization
摘要: 为了提高求解半定规划问题的运算效率,提出了一种新的求解半定规划的非单调信赖域算法。将半定规划的最优性条件转化为无约束优化问题,并构造无约束优化问题的信赖域子问题,修正信赖域半径的校正条件,当初始搜索点处于峡谷附近时仍能搜索到全局最优解。实验结果表明,对于小规模和中等规模的半定规划问题,该算法的迭代次数都比经典的内点算法少,运行速度快。
关键词: 半定规划, 信赖域算法, 非单调策略, 内点算法, 无约束优化
CLC Number:
TP301.6
YU Dongmei GAO Leifu. New algorithm for semidefinite programming[J]. Journal of Computer Applications, 2014, 34(1): 182-184.
于冬梅 高雷阜. 求解半定规划的新算法[J]. 计算机应用, 2014, 34(1): 182-184.
0 / Recommend
Add to citation manager EndNote|Ris|BibTeX
URL: https://www.joca.cn/EN/10.11772/j.issn.1001-9081.2014.01.0182
https://www.joca.cn/EN/Y2014/V34/I1/182