摘要
针对现有DS改进算法度量冲突程度方法的不准确与收敛速度慢的缺陷,提出证据集期望的概念和一种基于冲突系数和Pignistic概率距离的证据融合规则。用冲突系数和Pignistic概率距离的平均值表征证据间的冲突程度,通过计算每条证据的重要程度权值,进而求出证据集期望。用证据集期望作为现有证据集的反馈替换证据集中相关性最低的证据,以此修正证据集中可信度低的干扰证据,对修正证据集用DS规则进行融合。通过算例分析,验证了改进的组合规则能够排除干扰证据,降低冲突,并加快系统收敛速度。
Aiming at the shortcoming in the inaccurate measurement of the conflictive degree and slow convergence rate among the existing algorithm of DS evidence,the concept of expectation of evidence set and one rule of evidence fusion based on Pignistic probability distance are presented.The extent of conflict between two evidences is characterized by using the average value of collision coefficient and Pignistic probability distance.By calculating the weight of the importance degree of each evidence,the expectation of the evidence set is obtained.As the feedback of the existing evidence set,the expectation is used to replace the lowest correlation of evidence in evidence set.All evidence is combined in modified evidence set with DS rule.The numerical examples verify that the improve combination rule can identify and eliminate interference evidence to reduce conflict and accelerate the convergence rate of the system effectively.
出处
《桂林电子科技大学学报》
2017年第4期317-321,共5页
Journal of Guilin University of Electronic Technology
基金
国家自然科学基金(61362007)
广西信息科学实验中心基金(YB1407)
广西可信软件重点实验室基金(KX201414)
