计算机工程与应用 ›› 2014, Vol. 50 ›› Issue (24): 151-155.

• 数据库、数据挖掘、机器学习 • 上一篇    下一篇

DSm组合规则的改进及其在直觉模糊集上的拓展

唐长春,敖志刚,宫云祥,张康益,王有成   

  1. 解放军理工大学 野战工程学院,南京 210007
  • 出版日期:2014-12-15 发布日期:2014-12-12

Improvement of DSm combination rule and its development on intuitionistic fuzzy sets

TANG Changchun, AO Zhigang, GONG Yunxiang, ZHANG Kangyi, WANG Youcheng   

  1. College of Field Engineering, PLA University of Science and Technology, Nanjing 210007, China
  • Online:2014-12-15 Published:2014-12-12

摘要: DSmT在组合结果中保留了冲突焦元,一方面提高了组合冲突证据能力,另一方面造成了冲突信息的积累,促使分配给主焦元的信度不合理地减小,导致组合信息的不确定性增加,不利于决策。针对DSmT存在的缺陷,提出了一种新的基于冲突再分配的组合规则。为进一步提高DSmT处理冲突、模糊和不精确信息能力,研究了DSmT在直觉模糊集上的拓展,将传统的信度赋值函数替换为直觉模糊信度赋值映射,并讨论了直觉模糊拓展后的信度组合规则。数值实例验证了所提方法的有效性。

关键词: 信息融合, DSmT, 冲突再分配, 直觉模糊集

Abstract: Dezert-Smaradache Theory retains conflicting focal element in the combined result, which benefits the combination ability for conflict evidence, meanwhile, it results in the accumulation of conflicting information, and makes the composite reliability of main focal element reduce illogically. A new combination rule based on conflict redistribution is proposed to resolve the problem of DSmT. In order to further improve the DSmT performance in dealing with uncertain, highly conflicting and imprecise information, this paper researches the method to develop DSmT on the intuitionistic fuzzy sets. An intuitionistic fuzzy belief assignment mapping is defined to replace traditional belief assignment function, and the combination rule based on the new mapping is discussed in this paper. Numerical examples demonstrate the effectiveness of the proposed method.

Key words: information fusion, DSmT, conflict redistribution, intuitionistic fuzzy sets