摘要
对证据理论和直觉模糊集理论之间的本质联系进行分析,提出一种基于直觉模糊集改进的证据合成实用算法用于多传感器目标识别.根据直觉模糊集中隶属度和非隶属度的概念,对证据理论的可信度函数模型进行改进,提出了直觉模糊可信度分配函数模型并构造了相应的证据合成规则,以提高证据合成计算效率,使合成结果便于最终决策.通过与其他算法的对比实验验证了所提出算法的有效性.
Essential relationship between intuitionistic fuzzy sets(IFS) theory and evidence theory is discussed,and a practical evidence combination algorithm based on IFS is proposed for the multi-sensor target recognition application.The basic probability assignment function is modified based on the conceptions of membership degree and non-membership degree in IFS theory,and evidence combination algorithm is also modified accordingly in order to improve the computational efficiency and facilitate the final decision.Experiments comparing with other approximations show the effectiveness of the proposed algorithm.
出处
《控制与决策》
EI
CSCD
北大核心
2012年第11期1725-1728,1734,共5页
Control and Decision
基金
航天支撑技术基金项目(101.1.5)
关键词
信息融合
目标识别
证据理论
直觉模糊集
基本可信度函数
information fusion
target recognition
evidence theory
intuitionistic fuzzy sets
basic probability assignment function
作者简介
耿涛(1984-),男,博士生,从事系统工程、复杂系统建模的研究;
卢广山(1963-),男,研究员,博士生导师,从事系统工程等研究.