Abstract
With the increasing complexity of real decision-making problems, some experts try to use multi-scale information to express expert opinions in group decision-making problems. Facing the problem of group decision-making with multi-scale information, this paper attempts to explore a multi-scale group decision-making method consisting of two stages to provide theoretical support and methodological basis for establishing a multi-scale decision analysis system. In the first stage, we introduce a newly ranking decision-making approach based on a reflexive fuzzy alpha-neighborhood operator to deal with single-scale ranking problems with a single expert, which greatly enriches the ranking decision analysis theory. We also use a numerical example and experimental analysis to detect the stability and validity of the method. In the second stage, considering that the decision-making opinions of multiple experts may appear at the same time in the decision-making process, we propose the score labeling approach and the decision fusion approach to obtain the comprehensive decision-making result, which provides a feasible research idea for the comprehensive analysis of group decision-making results. Combining these two stages, a complete multi-scale group decision-making method in a multi-scale environment is described in detail, which can effectively deal with multi-scale group decision-making problems. Moreover, a series of simulation calculations are conducted to test the validity and stability of the proposed group decision-making method.