摘要
提出了一种新的融合模糊传感器数据的方法 .首先把模糊传感器输出的隶属函数作为对模糊命题的支持程度 ,之后引入了一个相似性测度对各个传感器输出的相似性进行度量 ,定义了一个相对一致度系数表示各个传感器被其他传感器所支持的程度 ,最后基于幂均算子实现了模糊传感器的数据融合 .与已有的方法相比 ,该方法综合考虑了各个传感器的输出信息 ,降低了融合过程中的信息损失 .通过一个目标识别的算例验证了该方法的有效性 .
This paper presented a novel method to fusion fuzzy sensor data based on a power average fusion operator. The fuzzy sensor's output was modeled as a membership function to describe the degree of support for a hypothesis. A similarity measure was introduced to determine the similarity between each sensor's outputs. Then a relative consensus degree coefficient was defined to represent each sensor's support degree by other sensors in the multi-sensor systems. The final fusion result was obtained by the power average operator. Compared with the other methods, this approach can decrease the information loss effectively. A numerical example was illustrated to show the fusion algorithm's efficiency.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2003年第8期1279-1281,1287,共4页
Journal of Shanghai Jiaotong University
关键词
模糊传感器
幂均融合算子
目标识别
fuzzy sensor
power average operator
object recognition