Dempster-Shafer theory of evidence based on BP and its application
Lingyan Liu · Journal of Naval University of Engineering · 2007
Demmpster-Shafer theory of evidence has good performance in dealing with uncertain information and has been used in many fields.However,basic probability assignment(BPA) is one of the key problems in its application and we still have not an objective approach for acquiring the propositions′ BPAs.These current approaches are mostly influenced by the experts′ knowledge and preference,and these obtained BPAs are difficult in describing the practical fact.In this paper,BP is applied to solving such a problem and integrated with D-S theory organically to deal with uncertain information fusion.The self-learning and adaptive capability as well as fault-tolerance may come into effect completely.A model for fault diagnosis using D-S theory of evidence based on BP is discussed in detailed.The fusion algorithm for combing all pieces of evidence is presented.Finally,an illustrative example of diesel engine fault diagnosis proves that the proposed approach is feasible.