A parametric linear relaxation algorithm for globally solving nonconvex quadratic programming

被引:15
|
作者
Jiao, Hongwei [1 ]
Liu, Sanyang [1 ]
Lu, Nan [1 ]
机构
[1] Xidian Univ, Sch Math & Stat, Xian 710071, Peoples R China
基金
中国国家自然科学基金;
关键词
Nonconvex quadratic programming; Global optimization; Parametric linearized technique; Pruning operation; BOUND ALGORITHM; APPROXIMATION ALGORITHMS; OPTIMIZATION ALGORITHM; BRANCH; CONSTRAINTS; CUT;
D O I
10.1016/j.amc.2014.11.032
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
In this article, we present a parametric linear relaxation algorithm for globally solving the nonconvex quadratic programming (NQP). In this algorithm, a new parametric linearized technique is proposed for generating parametric linear relaxation programming (PLRP) of the NQP, which can be used to determine the lower bound of global minimum value of the NQP. To improve the convergent speed of the proposed algorithm, a pruning operation is employed to compress the investigated region. By subdividing subsequently the initial domain and solving subsequently a series of parametric linear relaxation programming problems over the subdivided domain, the proposed algorithm is convergent to the global minimum of the NQP. Finally, an engineering problem for the design of heat exchanger network and some test examples are used to verify the effectiveness of the proposed algorithm. (C) 2014 Elsevier Inc. All rights reserved.
引用
收藏
页码:973 / 985
页数:13
相关论文
共 50 条