摘要
在对偶单纯形方法的基础上,提出了线性规划的目标函数最速递减算法。它避开求初始可行基或初始基,以目标函数全局快速递减作为选基准则,将选基过程与换基迭代合二为一,从而大大减少了迭代次数。数值算例显示了该算法的有效性和优越性。
On the basis of simplex method for dual, a fast decreasing algorithm of objective function is proposed. This method needn＇t find initial feasible basis or initial basis. The processes of choosing basis and exchanging basis are combined into one process by using the objective function globle fast decreasing criterion of choosing basis. And then the number of iterations is greatly reduced. Numerical results show its effectiveness and superiority.
出处
《运筹与管理》
CSCD
2005年第4期55-59,共5页
Operations Research and Management Science
基金
湖南省自然科学基金资助项目(03JJY3014)
湖南省教育厅科研基金资助项目(02C355)
关键词
线性规划
单纯形方法
对偶单纯形方法
目标函数最速递减算法
linear programing
simplex algorithm
dual simplex algorithm
fast decreasing algorithm of objective function
