带有界约束非凸二次规划问题的整体优化方法  

A Global Optimality Method for Solving the Nonconvex Quadratic Programming Problem with Additional Box Constraints

作　　者：高岳林[1] 徐成贤[1] 杨传胜[1] 

机构地区：[1]西安交通大学理学院

出　　处：《工程数学学报》2002年第1期99-103,共5页Chinese Journal of Engineering Mathematics

基　　金：国家自然科学基金资助项目 (199710 6 5 )

摘　　要：通过研究带有界约束非凸二次规划问题 ,给出了求解该问题的整体最优解的分枝定界方法及其收敛性 ;提出了定界的紧、松驰策略 ,把球约束二次规划问题作为子问题来确定原问题的整体最优值下界和上界 ,应用分枝定界方法达到了对原问题的求解。This paper researches the nonconvex quadratic programming problem with addition box constraints and gives a branch and bound algorithm of solving the problem as well as analyses the convergent property of the algorithm. This paper put forward bound tight or slack strategy, and regards ball constrained quadratic programming problem as sub problem in order to determine the upper and low bound of the global optimality values of the former problem, The former problem is solred by making use of branch and bound method.

关 键 词：非凸二次规划 分枝定界方法 整体优化 球约束 界约束 

分 类 号：O212.2[理学—概率论与数理统计]