Evidence Fusion Algorithm Based on Correction Conflict

In order to solve the shortcomings of traditional D_S evidence theory in dealing with conflict evidence,this paper presents a new measure. This method combines the evidence distance with the normalization constant to determine the degree of conflict between the evidences. The degree of conflict as the basis for judging the conflict evidence. Calculates the trust degree and the false degree of the evidence, sorts the trust degree and the false degree of the evidence respectively, then weights the conflict evidence. Finally, use the Dempster combination rule as evidence fusion. Experiments show that the algorithm not only judge the conflict evidence effectively, but also improve the convergence speed and precision.