摘要
本文在假定决策者的局部偏好结构为线性加权形式但权数未知的条件下,利用极大化极小规则,建立了一种多目标决策方法——迭代协调交互满意方法。通过迭代求得协调解,通过交互求得满意解。本文证明了算法的收敛性,并给出了决策者的最佳协调解、最佳协调权及最佳目标平均满足度等概念及其求法。最后通过数值例示说明了方法的可行性及适用性。
Under the assumption that the decision-maker's local preference structure is linear with the unknown weight, by means of the max-min rule in game theory, this paper develops amultiobjective decision-making method--Iteratively Compromising and Interactively Satisfyingmethod (ICIS). The compromising solution is obtained by iteration, and the satisfying solution is obtained by interaction. The convergence of ICIS method is proved and the optimal compromising weight, compromising solution and compromising average satisfying degree are given. Finally, the feasibility and applicability of ICIS method are demonstrated by a numerical example. <
出处
《系统工程与电子技术》
EI
CSCD
1992年第7期19-26,共8页
Systems Engineering and Electronics
关键词
多目标决策
决策支持系统
迭代解
Multiobjective decision making, Decision analysis, Satisfying solution, Compro- mising solution.
