The Study Of Relevant Evidence Fusion Method Based On The Separable Mass Funsion

In the research of evidence theory evidence fusion, study relevant evidence fusionis a very important direction, evidence theory is an uncertainty reasoning method,fusion of evidence is an important extension of evidence fusion, compared the twotreatment evidence. Evidence theory in different conditions of evidence synthesis whenthe emphasis is on evidence of the independence between the actual situation, and tocollect evidence often have some degree of correlation, this time with the theory ofevidence the results of synthesis will produce super estimation, that is no longer usesthe evidence theory. Aiming at the shortage of evidence theory, through the study ofrelevant evidence structure, combination method is given. Can be separated based onmass function. Finally the arbitrary structure mass function gives an approximationalgorithm, makes the evidence fusion becomes possible.First of all, the paper introduces relevant evidence fusion method of separable massfunctions based on the background and significance of the research, research status quoof evidence theory are outlined.Secondly, the thesis introduces the related knowledge of evidence theory andevidence synthesis, as well as the basic problem to be solved in this paper.Thirdly, aiming at the limitations of the evidence theory evidence evidenceindependence requirements, the synthesis method of separable mass functions based onthe relevant evidence, the identification of the relevant parts of the algorithm indifferent decision theory, the superiority of the evidence theory research evidencesynthesis is further verified by examples.Then, according to the evidence for a special class of separable mass functionsynthesis method. About any separable mass function is decomposed into a single setof simple mass function focal element decomposition rule. Finally, the relevant evidence synthesis in either mass functions, we first give theseparable approximation method of evidence function, then the synthesis of relevantevidence according to the separable mass function criterion. It resolves all kinds ofrelevant evidence synthesis.