Title :
An Effective Combination Rule When Evidence Conflict
Author :
Li, Fei ; Wang, Xiaofeng
Author_Institution :
Dept. of Comput. Sci., Shanghai Maritime Univ., Shanghai, China
Abstract :
Theory of evidence is a kind of effective treatment methods for uncertain information fusion. However there will be problems in the fusion results based on the classical D-S combination when the evidence serious conflict or completely conflict. This paper proposes a kind of effective combination of fusion conflict evidence. At first a distance function was introduced. Then the similarity of evidence was said by the similarity functions and through the normalization got the support of evidence as the weight of evidence. Finally to realize recycling information fusion by the average weight and classic combination rule.
Keywords :
inference mechanisms; sensor fusion; uncertainty handling; Dempster-Shafer evidence combination; combination rule; distance function; evidence similarity; evidence theory; evidence weight; fusion conflict evidence; information fusion; Cognition; Computer science; Convergence; Educational institutions; Electronic mail; Information science; Millimeter wave technology; combination rule; conflict evidence; similarity function;
Conference_Titel :
Intelligent Human-Machine Systems and Cybernetics (IHMSC), 2011 International Conference on
Conference_Location :
Zhejiang
Print_ISBN :
978-1-4577-0676-9
DOI :
10.1109/IHMSC.2011.73
