基于组合规则的证据合成方法.doc

基于组合规则的证据合成方法.doc基于组合规则的证据合成方法
摘要:针对传统Dempster?Shafer证据合成方法合成冲突证据时会出现相悖结论的问题,提出基于组合规则的证据合成方法。通过肯德尔等级相关系数得到证据间的相关程度,通过算法步骤顺序求解证据距离以及权重系数后,对证据进行概率重新分配,完成证据预处理过程。引入命题支持度到新的合成规则,完成再分配证据的最终合成。同时提出算法性能模型用于分析算法优劣。实例验证Y算法在合成冲突证据时具存最优结果,且和推理一致,同时也可以用于常规证据合成,算法具有通用性和普适性。
本文采集自网络,本站发布的论文均是优质论文,供学习和研究使用,文中立场与本网站无关,版权和著作权归原作者所有,如有不愿意被转载的情况,请通知我们删除己转载的信息,如果需耍分享, 请保留本段说明。
关键词:证据理论;冲突;权重;肯德尔等级相关系数;命题支持度
中图分类号:?34; TP391文献标识码:A文章编号:
10047373X (2017) 13?0122?04
Abstract : The conflict evidence bined with
traditional Dempster?Shafer bination method, which
may appear the problem of inconsistent conclusion, therefore
an bination method based bination rule is
proposed. The correlation degree among the evidences is obtained
according to Kendall rank correlation coefficient. The evidence
distance and weight coefficients are solved sequentially by
means of the steps of the algorithm to perform the probability
redistribution for evidences, and preprocess the evidence. The
proposition support degree is introduced into the new synthesis
rule for bination of the redistribution evidence. The
algorithm performance model is proposed to analyze the merits
of the algorithm. The example verification results show that the
algorithm has optimal result for the conflict evidence
synthesis, which is consistent with the reasoning result. The
algorithm can be used in conventional bination,
which is universa