摘要
采用变量消去法化二次规划问题 ,使用一个基于信赖域子问题的内点算法来获得其可行下降方向 ,提出了关于二次规划问题信赖域中可行下降的新算法 ,证明了算法具有全局收敛性 .计算实例表明 。
Quadratic programming is fundamental non-linear programming with constraints. A quadratic programming problem (1) is reduced to problem (2) with elimination method and an interor point algorithm is designed with sub-problem on the trust region. Using this algorithm, a feasible descent direction is obtained to get the minimum of step length for next approximation solution. A proof of global convergence of this algorithm and a computing example are provided.
出处
《武汉理工大学学报（交通科学与工程版）》
北大核心
2004年第5期732-735,共4页
Journal of Wuhan University of Technology（Transportation Science & Engineering）
基金
国家自然科学基金项目资助 (批准号 :70 3 710 63 )
