多源目标识别中冲突证据的融合策略研究

Research on Fusion Strategy of Conflict Evidence in Multi-Source Recognition Target Recognition

随着防空反导作战环境的日益复杂,快速准确地进行真假目标识别显得尤为重要。针对多源目标决策级识别中冲突证据融合问题,在综合分析现有改进证据理论方法的基础上,引入新的冲突度量系数,提出了一种通过在融合流程中增加智能判断环节,以冲突系数为判断依据,对具有不同冲突系数证据选择不同组合规则进行证据组合的融合方法,并给出了策略的流程图。最后进行了实例验证,结果表明新策略能对不同未知信度和不同识别框架冲突证据进行快速有效的融合。

Along with the more complex environment of air defense and antimissile, fast and exact recognition of the true and false targets is becoming more and more important. Based on the problem of conflict evidences fusion in multi source decision-making target recognition, a new conflict metric coeffi- cient is introduced on the basis of synthetically analyzed current improved evidences combination rules. The intelligent judge process is set in the fusion flow according to the conflict coefficient. The different combination rules are used on evidences with different coefficients; moreover, the strategy flow chart is put forward. The example is validated in the end, and the result proves that the conflict evidences of dif- ferent unknown faith rates and recognition frames can be fused by the new method.