Improved Evidence Theory Based on Partial Correction of Evidence Dissimilarity
Download as PDF
DOI: 10.23977/AICT2020030
Author(s)
Pin Wang, Xue-dong Xue, Wei Liu, Kan Cheng, Long-fei Jiang
Corresponding Author
ABSTRACT
Considering the difficulty of conflict evidences combination in target fusion recognition, a new improved approach based on partial correction of evidence dissimilarity is proposed from the perspective of evidence preprocessing. The proposed approach obtains judgment method of conflict evidences from the evidence dissimilarity and maximizes the support for non-conflicting evidences by use of the pretreatment of partial correction. By simulations, the results show that the proposed approach can combine the evidences with high conflict effectively and has better performance in recognition reliability, anti-interference and convergence speed.
KEYWORDS
Evidence theory; partial correction; evidence dissimilarity; target recognition