系统工程与电子技术 ›› 2025, Vol. 47 ›› Issue (6): 1909-1916.doi: 10.12305/j.issn.1001-506X.2025.06.19

• 系统工程 • 上一篇    下一篇

融合属性内部不确定和外部交互的多属性群体决策模型

王嘉丽, 江文奇   

  1. 南京理工大学经济管理学院, 江苏 南京 210094
  • 收稿日期:2024-03-08 出版日期:2025-06-25 发布日期:2025-07-09
  • 通讯作者: 江文奇
  • 作者简介:王嘉丽(1994—), 女, 博士研究生, 主要研究方向为多属性决策、模糊数据集
    江文奇(1975—), 男, 教授, 博士研究生导师, 博士, 主要研究方向为复杂决策
  • 基金资助:
    国家自然科学基金(71971117);国家自然科学基金(71671001);教育部人文社科基金(17YJA630035)

Multi-attribute group decision-making model integrating internal uncertainty and external interaction of attribute

Jiali WANG, Wenqi JIANG   

  1. School of Economics and Management, Nanjing University of Science and Technology, Nanjing 210094, China
  • Received:2024-03-08 Online:2025-06-25 Published:2025-07-09
  • Contact: Wenqi JIANG

摘要:

个性化个体语义体现不同决策者对语言术语理解的差异, 而二元语义在表征这种个性化个体化语义方面具有更大的优势。针对二元语义距离测度及其属性权重设计的难点, 系统研究基于二元语义距离测度的特性, 提出一种距离测度公式。首先, 基于证据理论和模糊测度分别设计属性内部不确定和外部交互作用下的双目标线性规划模型,以求解综合属性权重。接着, 基于备选方案的效用值矩阵和排序矩阵, 给出一种二元语义评估值的群决策求解方法。算例和对比分析验证所提出的方法可以有效处理属性的不确定信息和属性交互行为, 所提方法具有较强的鲁棒性, 使排序结果更加准确和稳定。

关键词: 属性权重, 距离测度, 个性化个体语义, 二元语义

Abstract:

Personalized individual semantics represent the different understandings for language terms by different decision-makers, and binary semantics have greater advantages in expressing personalized individual semantics. This paper focuses on the difficulties in designing binary semantic distance measurement and its attribute weights, systematically studying the characteristics of binary semantic distance measurement and proposing a distance measurement formula. Firstly, a dual objective linear programming model is designed based on evidence theory and fuzzy measure, respectively, for the internal uncertainty and external interaction of attributes to solve the comprehensive attribute weights. Secondly, based on the utility value matrix and sorting matrix of alternative solutions, a group decision solving method with binary semantic evaluation values is proposed. Numerical examples and comparative analysis verify that the proposed method can effectively handle the uncertain information and interaction behavior of attributes. The proposed method has strong robustness, making the sorting results more accurate and stable.

Key words: attribute weight, distance measurement, personalized individual semantics, binary semantics

中图分类号: