关于结构元理论的模糊多属性群决策.doc

摘  要  群决策作为决策理论的前沿科学 ,是解决群体决策问题的有效手段。多属性群决 策作为现代决策科学的重要组成部分，已经被广泛应用于政治、经济、军事等诸多 领域，它的特点是在多个属性多个方案的情况下，每个决策的参与者分别进行独立 的判断与决策，然后将各决策者的判断和决策的结果进行综合，从而达到群体决策 的效果。在决策过程中，由于客观事物的复杂性、不确定性以及人类思维的模糊性， 需要处理大量的模糊数据。因此，在模糊环境下对多属性群决策方法进行研究具有 重要的理论和现实意义。 模糊环境下的多属性决策问题通常可以转化为对模糊数的排序问题。本文通过 两种新的模糊数排序方法—模糊数排序的指数方法、基于质心和破碎度的模糊数排 序方法作为引子，引出对复杂的模糊多属性群决策问题的研究。 研究模糊多属性群决策问题，将模糊结构元理论引入决策过程中，可以简化传 统决策的复杂运算。本文以结构元理论为基础，给出了两种基于结构元理论的模糊 多属性群决策方法。第一种方法通过分析决策者评价的一致性程度对群体评价结果 的影响，结合决策者权重，将这种主观权重与客观权重集结为决策者的综合权重， 使得评价结果更符合实际；第二种方法将结构元理论同经典的 ELECTRE 方法结合， 弥补了使用 ELECTRE 方法时难以对模糊数进行排序、需要转换成确定系统的不足； 基于模糊结构元理论的模糊多属性群决策方法将模糊数的复杂运算转为同序单调 函数的运算，通过单调函数间的序关系来描述模糊数之间的序关系，从而简化了复 杂的决策过程。本文最后通过具体实例来说明这两种方法的有效性和合理性。 关键词：多属性群决策；模糊结构元；一致性分析；ELECTRE 方法；模糊数 - I - Abstract Group decision-making is an effective means as the forefront of science for decision making theory. Multi-attribute group decision making has been widely used in political, economic, military and other fields as an important component of decision science. In order to achieve the effect of group decision making, every decision makers judge and make decision independently and then integrated the results of judgments and decisions as the main features in the case of multiple properties under multiple programs. As the complexity of objective things, uncertainty and ambiguity of human thinking, the process of decision making need to deal with a lot of fuzzy data. Therefore, multi-attribute decision making research has important theoretical and practical significance in fuzzy environment. Fuzzy multi-attribute group decision making problem can be usually transformed into fuzzy numbers ranking problem. The power method for ranking fuzzy numbers and ranking fuzzy numbers based on centroid and fragmentation are proposed in order to elicit the research on fuzzy multi-attribute group decision making problem. In this paper, the structured element theory is introduced into fuzzy multi-attribute decision making in order to simplify the complex operations of