An Improved DS Evidence Combination Algorithm Applied to Information Fusion Algorithm Performance Evaluation

The DempsterShafer evidence theory is introduced to decide the optimal algorithm for information fusion management. Moreover, an improved evidence combination approach is proposed such that the overestimate generated from the combination of relevant evidence is suppressed. The relevance measurement function is generated and the basic probability assignment of the relevant evidences is adjusted. And then, the revised evidence is further combined. A numerical example provided shows the rationality and efficiency of the proposed approach.