一种多人有关联两层多目标决策的交互式算法

An lnteractive Algorithm for Two-level MultiobjectiveDecision Making with Several Decision Makers interconnected

本文首先建立了基于Ｓｔａｃｋｅｌｂｅｒｇ主从策略的多人有关联的两层多目标决策问题的数学模型，利用满意度和Ｋｕｈｎ－Ｔｕｃｋｅｒ条件把两层多目标规划问题转化为单层单目标非凸约束规划问题；并采用收敛外部逼近法求解此非凸约束规划问题的全局最优解；然后，通过分析人与决策人之间的交互，求得两层决策问题的满意解．

in the paper, a mathematical model is set up of two-level multiobjective decision making with several decision makers interconnected based on the Stackelberg leader-follower game, and with the help of satisfactoriness and Kuhn-Tucker condition, the two-level multiobjective programming problem is changed a singular-level, singular-objective, nonconvex constraint programming problem to which the convergent outer approximation method gives the global optimal solution afterwards. And then, the analyst and decision makers interact to achieve the preferred solution to the two-level decision making problem.