A reasoning method in Conditional Evidential Networks based on DSm-PCR5

To solve the problem that the evidence information on Dezert-Smarandache (DSm) model can not be effectively conditionally reasoned in multi-source heterogeneous a reasoning method in Conditional Evidential Network(ENC) based on DSm-Proportional Conflict Redistribution No.5(PCR5) is proposed. First, the conditional reasoning formula and the reasoning algorithm in Conditional Evidential Network based on DSm model is given. Then, the Disjunctive Rule of Combination(DRC) based on DSm-PCR5 is proposed and the Generalized Bayesian Theorem (GBT) for intersection sets of focal elements is obtained in the that the conditional mass assignments functions of focal in refinement of hyper-power set is known. Finally, the simulation e,periments results of situation assessment, the effectiveness of the proposed method is verified.