多目标决策的逼近方法（Ⅱ）：应用与数值分析  

APPROXIMATION METHOD FOR MULTIOBJECTIVE DECISION MAKING(Ⅱ):APPLICATIONS AND NUMERICAL ANALYSIS

作　　者：胡新生[1,2] 周济[1,2] 施保昌 陈王廷[1,2] 

机构地区：[1]华中理工大学CAD中心 [2]华中理工大学系统工程研究所

出　　处：《系统工程学报》1996年第3期11-20,共10页Journal of Systems Engineering

摘　　要：本文从数值计算的角度对求解多目标ｍｉｎｉｍａｘ问题的一种有效的逼近方法：极大熵方法进行了研究．对逼近问题用多目标决策的交互式可行方向法求解，用Ｃ语言编制了算法的计算程序并在３８６微机上对许多问题做了试算．数值结果表明：本文算法具有计算速度快、精度高、适用范围广且易于理解和使用等优点．We study a kind of efficient approximation method: maximum entropy method for multiobjective minimax problems from the point view of numerical computation. The approximating problem is solved by interactive feasible direction method which is coded by C language on microcomputer 386. Many test problems are computed by the code. The numerical test shows that the method here has the advantages of fast rate and high accuracy of computation, flexibility and that it can be easily understood and used.

关 键 词：多目标决策 逼近法 数值分析 非线性规划 

分 类 号：O225[理学—运筹学与控制论]