群决策中基于一致性强度的专家意见集结方法

Aggregation programming of experts' judgment information based on group consistency intensity in group decision-making

针对群决策问题,提出了一种基于群体一致性强度的虚拟专家意见逐步集结规划的方法。首先将两个专家的偏好信息集结为序关系值向量,将该次集结结果作为一个'虚拟专家'给出的判断信息,然后吸收下一个专家的意见成为新的虚拟专家,依此类推,直到最后一个专家参与决策。引入虚拟专家的群体一致性强度保证群体结果趋向一致,其中一致性强度和虚拟专家的权重依赖于所包含的个体专家的人数。个体人数越多一致性强度和权重就越大,从而体现了少数服从多数的原则。该方法将多专家的复杂群决策问题转化为低复杂度的多阶段二人决策问题,并保证了群体结果的一致性。最后,通过一个算例验证了方法的有效性。

An aggregation programming group decision-making method by integrating virtual experts＇ judgment based on group consistency intensity is proposed. First, two experts＇ judgment information is integrated as preference ordering, and the result of this integration is regarded as the new judgment given by a ＂virtual expert＂. Then, this virtual expert absorbs the information of a new expert as the next virtual expert. By the same means, another expert is absorbed in the previous virtual expert until the last. During this process, group consistency intensity is introduced to make sure the gradual eonvergence of the ultimate outcome. The group consistency intensity and the weight of every virtual expert depend on the individual number of experts included in the virtual expert. The larger the number of individual is, the greater the weight of virtual expert is, and thus the rule of majority is reflected. This method transforms the complicated multiple experts＇ group decision question into a low complexity decision question of two people, and guarantees the eonsisteney of colony＇s result. An example shows the validity of the method.