基于支持度的限定记忆二次融合算法  

Limited memory second fusion algorithm based on support degree

作　　者：张建业[1] 王占磊[2] 张鹏[2] 陈芳[3] 刘博宁[2] 

机构地区：[1]空军工程大学科研部,西安710051 [2]空军工程大学工程学院,西安710038 [3]空军工程大学训练部,西安710051

出　　处：《计算机应用研究》2012年第4期1338-1340,共3页Application Research of Computers

基　　金：国防预研基金资助项目(9140A27020308JB3201);航空科学基金资助项目(20100818017)

摘　　要：在先验知识未知的情况下对多传感器信息进行融合,提出了基于支持度的限定记忆二次融合算法。为了保证融合权重分配的合理性,算法利用绝对距离度量传感器观测值之间的相互支持程度,并引入限定记忆控制项来防止陈旧信息引起的数据饱和。为了进一步保证故障发生时的融合效果,在一次融合的基础上,引入灰关联度对量测数据进行优势分析,并对分析调整后的数据进行二次融合。仿真结果表明,该算法可进一步提高融合精度,具有较强的容错和抗干扰能力。In the case of any prior knowledge was unknown,this paper proposed the limited memory second fusion algorithm based on support degree.To guarantee the reasonability of weight coefficient assignment,it used absolute distance to measure the mutual support degree of observation value of multi-sensors,and proposed a limited memory fusion to avoid data saturation when the old measurement information increasing.In order to improve the fusion effect when failure occurred,on the basis of first fusion,it introduced the ray relationship degree to preponderantly analyse measurements,then the second fusion of processed data was done.The simulation result shows the new algorithm can further improve the accuracy of fusion,and make the fault tolerance and anti-jamming ability strengthen.

关 键 词：支持度 限定记忆 灰关联度 二次融合 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]