A Conflict Evidence Fusion Method Based on Vector Metrics

Abstract

The invention discloses a conflicting evidence fusion method based on vector measurement, which can realize the measurement of the degree of conflict between evidences in multi-source information fusion, determine the weight factor of the fused evidence through the measurement of the degree of conflict between the evidences, and evaluate the fused evidence After correction, reliable decision-making results of target recognition can be obtained by fusion of Dempster combination rules. Compared with the traditional algorithm, the scheme of the present invention starts from the perspective of the evidence vector, comprehensively considers the degree of conflict between the evidence vectors and the degree of similarity to jointly measure the degree of conflict between the evidence, and determines the weight factor of the fusion evidence through the evidence conflict degree factor, and performs a calculation on the evidence. After the correction, the Dempster combination rule is used to fuse the revised evidence one by one to obtain a reasonable decision result, which can be well applied in the field of target recognition and has important theoretical significance and application value.

Description

technical field [0001] The invention relates to a multi-source information fusion method, in particular to a vector measurement-based conflict evidence fusion method. Background technique [0002] The development of sensor technology provides hardware support for military and civilian fields to obtain rich information. Because sensors are affected by external interference or human factors in complex environments, the output identification target information is uncertain and fuzzy. It may even be contradictory. If the problem of high-conflict evidence fusion cannot be effectively dealt with, the actual application system will not be able to make effective decisions, which will greatly affect the decision-making performance of the fusion system. However, traditional algorithms often take into account the influence of the difference degree of evidence vectors. , failed to consider other decision-making factors, so there is a certain error in the judgment of target recognition. ...

AI Technical Summary

Problems solved by technology

[0002] The development of sensor technology provides hardware support for military and civilian fields to obtain rich information. Because sensors are affected by external interference or human factors in complex environments, the output identification target information is uncertain and fuzzy. It may even be contradictory. If the problem of high-conflict evidence fusion cannot be effectively dealt with, the actual application system will not be able to make effective decisions, which will greatly affect the decision-making performance of the fusion system, and traditional algorithms often take into account the influence of evidence vector differences. , failed to consider other decision-making factors, so there is a certain error in the judgment of target recognition

Images