Research of combination rule based on evidence distance and local conflict distribution

Jun Wei Li, Yong Mei Cheng, Shao Wu Zhang, Yan Liang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

Dempster-Shafer Theory(DST) is an effective method for dealing with multi-sensor information. However, the results may be contrary to intuition when dealing with the highly conflicting evidence. In order to efficiently overcome the shortcomings of DST, a new improved DST combination rule based on evidence distance and local conflict distribution strategy is proposed. Firstly, the mutual support degree of every evidence is obtained by the evidence distance, and normalize the support degree to the relative credibility of the evidence; the generate reason of local conflict is analyzed, the conflicting information is only distributed to those focal elements which contribute to the conflict. It through calculating local conflict, and the distributed value depends on the credibility and similarity degree of the focal elements between the evidence. The results of numerical examples prove that the new combination rule enhances the reliability and rationality of the evidence combination results, and effectively solves the combination of highly conflicting evidence.

Original languageEnglish
Title of host publication2010 Chinese Control and Decision Conference, CCDC 2010
Pages2138-2142
Number of pages5
DOIs
StatePublished - 2010
Event2010 Chinese Control and Decision Conference, CCDC 2010 - Xuzhou, China
Duration: 26 May 201028 May 2010

Publication series

Name2010 Chinese Control and Decision Conference, CCDC 2010

Conference

Conference2010 Chinese Control and Decision Conference, CCDC 2010
Country/TerritoryChina
CityXuzhou
Period26/05/1028/05/10

Keywords

  • Dempster's rule
  • Evidence distance
  • Information fusion
  • Local conflict
  • Weight coefficient

Fingerprint

Dive into the research topics of 'Research of combination rule based on evidence distance and local conflict distribution'. Together they form a unique fingerprint.

Cite this