电子学报 ›› 2018, Vol. 46 ›› Issue (11): 2738-2745.DOI: 10.3969/j.issn.0372-2112.2018.11.022

• 学术论文 • 上一篇    下一篇

互补证据组合规则

孙贵东, 关欣, 衣晓, 赵静   

  1. 海军航空大学, 山东烟台 264001
  • 收稿日期:2017-11-20 修回日期:2018-03-24 出版日期:2018-11-25 发布日期:2018-11-25
  • 通讯作者: 关欣
  • 作者简介:孙贵东 男,1989年3月出生,山东荣成人,博士研究生.主要研究方向为智能数据挖掘,模糊识别与决策.E-mail:sdwhsgd@163.com
  • 基金资助:
    国防科技卓越青年科技基金;国家自然科学基金重大研究计划(No.91538201);国家自然科学基金面上项目(No.61671463,No.61571454);泰山学者工程专项经费(No.ts201712072);山东省自然科学基金(No.ZR2017BG014)

Complementary Combination Rule for Evidence

SUN Gui-dong, GUAN Xin, YI Xiao, ZHAO Jing   

  1. Naval Aviation University, Yantai, Shandong 264001, China
  • Received:2017-11-20 Revised:2018-03-24 Online:2018-11-25 Published:2018-11-25

摘要: 针对Dempster组合规则"一票否决"和现有修正方法主观分配冲突信息的缺点,提出一种全新的证据组合方法.新方法直接利用定义的互补证据进行融合,不再采用Dempster组合规则中的纯乘性策略组合证据,成功解决了"一票否决"问题.并将冲突信息转化为证据权重融入组合规则进行直接融合,不再主观分配其信度.新的组合规则是一类广义的加性集成算子,既满足交换律又满足结合律,既能够直接组合单类证据,也能够基于TBM组合复合类证据.通过多传感器融合识别算例,验证和对比分析了新的组合规则的优势.

关键词: 证据理论, 组合规则, 互补证据, 冲突, 信度转换模型

Abstract: A novel combination rule for evidence is proposed to deal with the "veto problem" in Dempster rule and the subjectively assigning the conflict information in the existing modified methods.This novel method combines the defined complementary evidence directly instead of using the pure Dempster multiplicative strategy,which successfully solves the "veto problem".It transfers the conflict to the evidence weight which can be processed directly with the novel combination rule instead of assigning the conflict subjectively.It is regarded as a generalized additive aggregation operator,which not only satisfies the commutative law and associative law,but also can combine the single focus elements directly and compound focus elements based on TBM.With the help of a multi-sensor fusion recognition example,the superiority of the proposed combination rule is illustrated and compared in detail.

Key words: evidence theory, combination rule, complementary evidence, conflict, the transferable belief model(TBM)

中图分类号: