The research on conflict redistribution of evidential reasoning

Abstract

The evidential reasoning theory is often used to process the decision information in decision support. The evidential reasoning is qualified in dealing with uncertain information. However, its combination rules always have some limitations in processing conflict information. In order to solve the conflict problem in the combination of multi-source information, this paper analyzed the conflict solutions of different combination rules, and compared them from the standpoint about the translation of evidence conflict. Meanwhile, the paper proposed a new conflict redistribution rule. This rule is based on the proportional conflict redistribution (PCR) rule and redistributes conflict masses in the proportion of the nth power of the conflict element's belief assignment. The simulated experimental results have proved its validity. The researches on combination rules can also be applicable to other belief-based uncertainty theories.