符号几何规划问题的求解新方法  

A New Method for Solution of Signomial Geometric Programming

作　　者：靳利[1] 

机构地区：[1]河南机电高等专科学校基础部,河南新乡453000

出　　处：《河南机电高等专科学校学报》2014年第2期48-50,共3页Journal of Henan Mechanical and Electrical Engineering College

摘　　要：对符号几何规划问题,给出一种求其全局最优解的分支定界算法。利用不同的凸化方法将算法中关键的定下界操作转化为易于求解的凸规划问题。最后给出算法的收敛性证明和数值实验结果。A global optimization algorithm is presented for solving signomial geometric programming. The method is based on the branch and bound scheme and employs different convex method to obtain the lower bounds. As a result, the lower bounding subproblems during the algorithm search are all convex programming problems. Convergence of the algorithm is proven and the numerical results are given to show the feasibility of the proposed algorithm.

关 键 词：符号几何规划 凸化方法 分支定界 

分 类 号：O221.2[理学—运筹学与控制论]