证据距离的选取对冲突证据组合影响的研究

A Research on the Influence of Evidence Distance Selection on Combination of Conflict Evidences

为了研究证据距离的选取对冲突证据组合的影响,提出一种将不同的证据距离度量方式应用在加权证据组合中的对比方法。首先采用不同的证据距离度量方式得到证据相似度,依据证据相似度求得证据可靠度,之后用证据可靠度对证据进行加权求和得到平均证据,最终对平均证据进行证据组合,对比在不同的证据距离度量方式下上述证据组合过程的合理性,得到适合加权证据组合中权重生成的距离度量方式。采用4个算例,对不同的证据距离度量方式下求得的证据可靠度、平均证据和证据组合结果进行分析比较,实验结果表明,与其他证据距离相比,利用该方法选出的证据距离能够更有效地体现证据之间的差异性,所求取的证据可靠度与实际情况更为贴合。

A comparative method is proposed to study the influence of evidence distance selection on combination of conflict evidences and the method applies different evidence distance measurements to combination of weighted evidences. Firstly,different evidence distance measurements are used to get similarities of evidences.Then evidence credibilities are calculated using evidence similarities and the averages of evidences are calculated by summing up weighted evidences using the credibilities as weights.Finally,resulting averages are combined.The rationality of these combinations of evidences in different evidence distance measurements is compared to obtain a suitable weights generation in weighted evidence combination.Numerical examples are given to analyze and to compare evidence credibilities,averages of evidences and combination results of evidences.Experimental results and comparisons with other evidence distance measurements show that the evidence distance measurements selected by the proposed method reflect the difference between the evidences more effectively,and the calculated evidence credibilities are more consistent with the actual situation.