基于马田系统和φ_s转换的模糊积分多属性决策方法

Fuzzy integral multi-attribute decision making method based on Mahalanobis-Taguchi system and φ_s transformation

在实际决策问题中,决策属性间往往存在一定的交互作用,而传统决策方法并不能有效处理。针对这种情况提出了一种基于马田系统(Mahalanobis-Taguchi system,MTS)和φs转换的模糊积分多属性决策方法。该方法针对φs转换法利用属性权重确定λ模糊测度存在的问题,提出利用Shapley值代替属性权重来确定λ模糊测度,同时提出了一种基于马田系统的Shapley值测度方法,并给出了合理性分析。最后通过实例分析了不同交互度对决策结果的敏感性,并验证了利用Shapley值确定的λ模糊测度更有利于决策。

For actual decision making problems, the traditional decision making methods can not effectively deal with the interaction between attributes. To solve this problem, a fuzzy integral multiattribute decision making method based on Mahalanobis-Taguchi system （MTS） and φs transformation is proposed. In this method,to overcome the problem of the λ fuzzy measure identification method based on φs transformation and the weights of attributes, this paper uses the Shapley value instead of weight to determine λ fuzzy measure. At the same time, this paper uses the Shapley value identification method based on Mahalanobis-Taguchi system, and the rationality of the identification method is analyzed. An illustrative example is given to analyze the sensitivity of the decision results based on different interaction degrees and verify the inference that the Shapley value is better than weight for decision making.