计算机工程与应用 ›› 2015, Vol. 51 ›› Issue (22): 134-140.

• 数据库、数据挖掘、机器学习 • 上一篇    下一篇

广义犹豫正态模糊信息集成及其多属性群决策

马庆功   

  1. 常州大学 怀德学院,江苏 常州 213016
  • 出版日期:2015-11-15 发布日期:2015-11-16

Generalized hesitant normal fuzzy information aggregation and their application to multi-attribute group decision making

MA Qinggong   

  1. Changzhou University Huaide College, Changzhou, Jiangsu 213016, China
  • Online:2015-11-15 Published:2015-11-16

摘要: 定义了犹豫正态模糊元及其运算法则、得分函数、Euclidean距离等概念;提出了广义犹豫正态模糊有序加权平均算子,并研究其性质,该算子不仅尽可能多地保留决策者的偏好信息,还可依据决策者的主观意愿选择不同的参数和属性权重,使得决策结果达到决策者的期望值;紧接着对属性权重和算子参数赋予不同的数值,获取广义犹豫正态模糊有序加权平均算子的若干种特殊算子,并探讨两个常用算子的大小关系;针对属性权重完全未知的多属性群决策问题,构建一种基于广义犹豫正态模糊有序加权平均算子的群决策方法。该方法利用同一属性下所有方案属性值间的距离求得最优权重,然后将同一方案下各属性值集结成为综合属性值,进而得到方案优劣排序。通过实例分析说明该方法的可行性和有效性。

关键词: 犹豫正态模糊元, 多属性群决策, 信息集成算子, 计算机网络系统

Abstract: Hesitant normal fuzzy elements(HNFEs) as well as their operational laws, score functions and Euclidean distance are defined. Then, the generalized hesitant normal fuzzy ordered weighted averaging(GHNFOWA) operator is proposed and some desirable properties of the GHNFOWA operator are studied. The GHNFOWA operator not only preserves the decision maker’s preference information as much as possible, but also the values of the parameter and attribute weights can changes on the base on decision makers’ attitude to make the results fix the expected values of decision makers. Furthermore, some special cases of the GHNFOWA operator are given when the weight vector?and?operator?parameter takes different values, and the relationship between two common operators is studied. Finally, for multi-attribute decision making problems with the information of attribute weights is completely unknown, a method based on the GHNFOWA operator is investigated. The optimal weights are calculated by the distances of each alternative under an attribute, and then aggregate all the attribute values into the overall attribute values, which is followed by the ranking of the alternative. An example is given to demonstrate the developed method is practicality and effectiveness.

Key words: hesitant normal fuzzy elements, multi-attribute group decision making, information aggregation operator, computer network systems