冲突证据的相关系数度量方法

Correlation coefficient measurement for conflict evidence

复杂战场环境增加了侦察信息处理的不确定性,基于信度函数研究不确定信息尤其是冲突证据度量问题,提出用归一化的证据相关作为冲突证据度量的相关系数,针对现有冲突证据度量方法未分清证据单类命题与多类命题的缺点,分别定义并核关系矩阵与核关系矩阵修正并核相关与核相关,得到修正后的冲突证据度量。结合经典的冲突证据算例验证了该冲突证据相关系数度量方法的有效性。

The complex battlefield environment increases the uncertainty of reconnaissance information processing,so the uncertain information processing is investigated,especially the conflict evidence measurement,based on the belief functions.The normalized evidence correlation is defined as the correlation coefficient measurement for conflict evidence.In order to solve the drawback that the existing conflict evidence measurement methods do not distinguish the single set and multiple sets,the union core and core relational matrix are defined to modify the union core and core correlation respectively and further obtain the modified conflict evidence measurement.Finally,the efficiency of the proposed correlation coefficient measurement for conflict evidence is illustrated in the classical simulation example.