摘要
为解决传统基于DS理论的融合算法在处理高冲突证据时,存在将所有证据都视为高冲突证据统一进行修正的问题,提出一种处理证据高冲突性的改进算法。判断证据的冲突程度,只对冲突性较大的证据进行修正,在保证原始证据真实性的基础上对高冲突证据源进行修正,用DS合成法则进行融合,并通过实验验证。实验结果表明:该算法具有良好的收敛性和可行性,与其他算法相比,可在一定程度上提高合成过程中证据的可靠性。
In order to solve the problem that the traditional fusion algorithm based on DS theory treats all the evidences as high conflict evidences in the process of handling them, this paper proposes an improved algorithm to deal with the high conflict of evidences. We judge the conflicting degrees of evidences firstly, then correcting the high conflict of evidences to reduce them based on the authenticity of original evidences and using DS synthesis rule to fuse the evidences, and then verifying the rationality through the experiment. The results of experiment show that the algorithm we proposed has good convergence and feasibility. Compared with other algorithms, the proposed algorithm can improve the reliability of evidence to some extent in the synthesis process.
引文
[1]杜晶,杨叶,王青,等.基于证据的可信软件过程评估方法[J].计算机科学与探索,2011,5(6):501-512.
[2]杨善林,丁帅,褚伟.一种基于效用和证据理论的可信软件评估方法[J].计算机研究与发展,2009,46(7):1152-1159.
[3]丁帅,鲁付俊,杨善林,等.一种需求驱动的软件可信性评估及演化模型[J].计算机研究与发展,2011,48(4):647-655.
[4]WANG B,ZHOU X,YANG G,et al.DS Theory-Based Software Trustworthiness Classification Assessment[C]//Symposia and Workshops on Ubiquitous,Autonomic and Trusted Computing.IEEE Computer Society,2010:434-438.
[5]MA Y J,CHEN Y X.An Attributes-based Allocation Approach of Software Trustworthy Degrees[C].2015IEEEE International Conference on Software Quality,Reliability and Security-Comoanion,2015.
[6]汪永伟,刘育楠,杨英杰,等.D-S证据理论中冲突性处理新方法[J].计算机工程与设计,2013,34(12):4316-4320.
[7]YANG J B.Rule and utility based evidential reasoning approach for multi-attribute decision analysis under uncertainties[J].European Journal of Operational Research,2001,131(1):31-61.
[8]孙媛,赵建军,周源.基于软计算技术的军用软件可靠性预测模型研究[J].兵工自动化,2017,36(2):56-60.
[9]张卫祥,刘文红,吴欣.软件可信性定量评估[M].北京:清华大学出版社,2015:140-141.
[10]朱友清,周石琳,邹焕新.基于相容系数的冲突证据合成方法及评价准则[J].系统工程与电子技术,2014,36(6):1118-1123.
[11]丁帅.软件可信性评估模型及其优化方法研究[D].合肥:合肥工业大学,2011:33-34.