随机型多目标决策的新方法

 设计了一种多目标随机型决策分析新方法——交互式参考目标满足度与可行度法(IRGSD-FD),讨论了随机型决策分析的希望水平原则、解及替代模型的定义和确定性等价等问题,并对IRGSD-FD的应用结果进行了讨论。

A NEW METHOD FOR MULTI-OBJECTIVE STOCHASTIC DECISION MAKING AND ITS APPLICATIONS

This paper is devoted to studying the mulii-objective stochastic decision making problem. A multi-objective stochastic decision making analysis method-Interactive ReferenceGoal Satisfied Degree and Feasible Degree (IRGSD-FD) method is developed, with the aspiration level principle for multi objective stochastic decision making analysis, the definitions of the solutions and surrogate models, and determinative equivalence problems, etc., being, discussed. The application results for IRGSD FD method are analysed, with Some useful results being obtained. The implementation of IRGSD FD method for stochastic decision making includes eight steps: (1) obtainment of the decision maker's initial reference goal level; (2) generation of the initial surrogate objective function and initial efficient solution; (3) evaluation of multiple attribute utility function of the decision maker for satisfied degree and feasible degrees (4) redefination of the sarrogate objective function; (5) generation of the efficient solution to be selected; (6) interaction with the decision maker; (7) test for compatibility; (8) generation of the new surrogate objective function and solution space.