基于粗糙D-S理论的身份融合方法  

Identity fusion method based on rough sets and D-S theory

作　　者：路艳丽[1] 雷英杰[1] 王晶晶[1] 

机构地区：[1]空军工程大学导弹学院,陕西三原713800

出　　处：《系统工程与电子技术》2007年第10期1749-1752,共4页Systems Engineering and Electronics

基　　金：国家自然科学基金(60773209);陕西省自然科学基金(2006F18)资助课题

摘　　要：综合运用粗糙集方法与D-S证据理论可有效解决身份融合中的一些复杂问题。针对多传感器身份融合中D-S理论获取基本概率赋值存在主观性的问题,提出了一种新的基于粗糙集理论的证据获取和基本概率赋值客观确定方法;利用粗糙集中属性的重要性消除冗余传感器,减少了证据组合的冲突现象。在此基础上,给出了基于粗糙D-S理论的身份融合方法,最后以具体算例验证了方法的有效性。Integrating rough sets and D-S evidence theory can efficiently solve some complex problems in identity fusion. First, to the question of subjectivity in getting basic probability assignment of DS theory for multi-sensors identity fusion, a new method of getting evidences and objective BPA based on Rough Sets theory is proposed. Then, conflict evidences are reduced by using the attribute weightiness of Rough Sets to eliminate redundant sensors. An identity fusion method based on Rough Sets and D-S theory is exposed with the foregoing research. Finally, the validity of the developed method is verified with a particular instance.

关 键 词：信息融合 粗糙集 证据理论 传感器 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]