基于灰关联度系数的证据组合方法  

Evidence Combination Method Based on Degree of Gray Relationship

作　　者：卢彦龙[1] 王拓[1] 徐宏[1] 

机构地区：[1]西安交通大学电子与信息工程学院,西安710049

出　　处：《计算机与数字工程》2017年第1期53-57,共5页Computer & Digital Engineering

基　　金：国家自然科学基金(编号:61071217);江苏省自然科学基金(编号:Bk0141216)资助

摘　　要：分析了D-S方法及其改进方法的不足,基于灰关联度系数提出了一种新的冲突证据处理算法。首先,以灰关联系数作为衡量标准,把每条证据当做离散的点,求出证据体集合中其它证据与它的相似程度。然后把相似程度进行归一化处理得到相似度,该相似度表示的是该证据的重要程度。把该相似度当做每条证据体的权重,对证据体加权合成新的证据体模型。最后用新证据体融合得到结果。通过算例表明:该方法与其他改进方法比较,具有收敛速度快,计算方法简单,易于工程实现的优点。In this paper,the shortage of D-S evidence theory as well as its modification are analyzed,then a new combination method based on degree of gray relationship is proposed.The method uses degree of gray relationship as evaluation criteria,then regards each piece of evidence as discrete points and computes the similarity between every two evidences in evidence collection.Then the obtained similarity which represents the importance of evidences is normalized.The similarity obtained above is regarded as the weight of every evidence and generate new evidence model.Finally fusion result is got by means of evidence model.The numerical examples show that the method proposed in this paper compared to other methods has following advantages,including fast convergent rate,simple calculation method,easy realization in a project.

关 键 词：D-S证据理论 证据冲突 灰色关联度 信息融合 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]