论文部分内容阅读
为有效克服证据理论(DST)组合规则存在的不足,提出一种基于局部冲突分配策略的DST改进算法.首先根据证据距离获得各个证据的相互支持度,并将支持度归一化为证据的相对可信度;然后通过分析局部冲突产生的原因,将冲突信息只分配给产生冲突的焦元,且分配给焦元的信息取决于证据之间焦元的信任度和相似度.实验结果表明,所提出的算法提高了证据合成结果的可靠性与合理性,可有效地解决高冲突证据融合的问题.
In order to overcome the shortcomings of DST combination rules effectively, this paper proposes a modified DST algorithm based on the local conflict distribution strategy. First, the mutual support of each evidence is obtained according to the evidence distance, and the support is normalized to the relative evidence Then, by analyzing the causes of local conflicts, the conflict information is only allocated to the focal units that generate conflicts, and the information assigned to the focal units depends on the trust and similarity of the focal units between the evidences.The experimental results show that, The proposed algorithm improves the reliability and rationality of the results of evidence synthesis and can effectively solve the problem of high-conflict evidence fusion.