Strong conflicting evidences fusion based on belief interval distance measurement

To suppress the counterintuitive results which are caused by the combination of strong conflicting bodies of evidence, we proposed a modified evidence combination. First of all, we found out that the belief interval on the evidence theory had shown the disadvantages of uncertainty of the evidence expression. And based on the differences of belief interval of all the elements in the identification frame between two bodies of evidence, we constructed a new distance measure to represent the conflict between bodies, and theoretically proved its rationality. Secondly, based on the measured distance result, we calculated the weight of multiple bodies of evidence participating in the fusion. And finally, the integration of conflicting evidence was resolved through fusion of the weighted average evidence. Numerical experiments show that when the conflict is weak, the new method is as same as the other existing advanced calculations, or similar with them. And when the conflict is strong, it can overcome the interference of strong conflicting evidences effectively and conform the results of intuitive cognition. Besides, its convergence speed is much faster.