摘要
将无约束正定式几何规划问题转化为等价的凸规划问题,利用目标函数的凸性,给出了一种在强Wolfe线搜索下的共轭梯度算法,并证明了其全局收敛性。
In this paper, we transformed the unconstrained positive geometric programming problem into the equally convex programming problem, and using the convex quality of the objective function, we gave a conjugate gradient method under the strong Wolfe linear search and proved its global convergence.
出处
《河南机电高等专科学校学报》
CAS
2008年第2期100-102,共3页
Journal of Henan Mechanical and Electrical Engineering College
基金
国家自然科学基金资助项目(106710570)
关键词
几何规划
凸规划
共轭梯度法
全局收敛性
geometric programming
convex programming
conjugate gradient method
global convergence