基于证据距离与不确定度的证据组合方法(英文)  

Weighted evidence combination based on distance of evidence and uncertainty measure

作　　者：韩德强[1] 邓勇[2,3] 韩崇昭[1] 侯志强[4] 

机构地区：[1]西安交通大学智能网络与网络安全教育部重点实验室机械制造系统工程国家重点实验室电信学院综合自动化研究所,陕西西安710049 [2]上海交通大学电子信息学院,上海200240 [3]西南大学计算机与信息科学学院,重庆400715 [4]空军工程大学电讯工程学院,陕西西安710077

出　　处：《红外与毫米波学报》2011年第5期396-400,468,共6页Journal of Infrared and Millimeter Waves

基　　金：the Grant for State Key Program for Basic Research of China(973)(2007CB311006);the Program for New Century Excellent Talents in University,Grant No.NCET-08-0345;Shanghai Rising-Star Program grant No.09QA1402900;Foundation for Innovative Research Groups of the National Natural Science Foundation of China(60921003);the National Natural Science Foundation of China(60805015);Research Fund of Shaanxi Key Laboratory of Electronic Information System Integration(No.201101Y17);China Postdoctoral Science Foundation(No.20100481337)

摘　　要：Dempster-Shafer证据理论是信息融合领域中的一种重要的理论与方法.然而在实际应用中,Dempster证据组合规则无法有效处理高冲突证据组合问题,往往引发反直观结果.针对这一问题,提出一种新的加权证据组合方法.该方法同时利用证据距离和证据不确定度来生成权重进而修正待征组合证据,并取得合理的组合结果.实验结果表明所提方法具有更快的收敛速度,能有效应对高冲突证据组合问题.Dempster-Shafer evidence theory is an important tool in the fields of information fusion.However Dempster＇s rule of combination cannot efficiently handle highly conflicting evidence combination for it can arouse counter-intuitive behaviors.To deal with such a problem,a novel weighted average evidence combination approach is proposed.Not only the distance of evidence but also the uncertainty measure is utilized to determine the weights of the bodies of evidence.Based on the weighted averaged BOE and Dempster＇s rule of combination,the rational combination results can be obtained.The experimental results show that the method proposed can effectively handle conflicting evidence combination with better convergence.

关 键 词：证据理论 传感器数据融合 距离函数 不确定度 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]