A new algorithm for conflicting evidence based on gray correlation

Dempster-Shafer (D-S) evidence theory is widely applied to many fields of information fusion. However, illogical results may be obtained by the classical Dempster combination rule when collected evidences are highly in conflict with each other. To suppress the counterintuitive results, this paper introduces the concept of gray correlation theory to calculate the correlation between the bodies of evidence and proposes a new algorithm to generate the discounting coefficients. Firstly, calculate the gray correlation degree between the bodies of evidence to obtain the degree of each evidence supported by other evidence. Secondly, obtain the discounting coefficients with a BUM function. Finally, modify the multi-source evidence by the discounting coefficient method before the combination based on Dempster's rule. A numerical example shows the efficiency and rationality of the proposed approach. © 2013 ISSN 1881-803X.