摘要
在分布式多传感器数据融合系统中 ,各单传感器类型不同情况下的融合问题是一类普遍的问题 .本文从分层融合算法的推导过程提出了一种解决这一类问题的方法 ,包括融合状态设定方法、对单传感器滤波结果的状态及其相应方差阵的扩维方法、扩维后的融合与预测方法 .同时明确了这种扩维融合算法的实质 .
In distributed fusion system,the problem that sensors of the system are of different type is a widespread problem.In this paper,from the process of hierarchical fusion,a resolution of fusion state definition,dimension extension of filter state vector from a single sensor as well as its covariance matrix,trace fusion and prediction after extension is introduced,followed by the essence of such kind of problem.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2000年第6期4-7,共4页
Acta Electronica Sinica
关键词
分层融合
多传感器
数据融合
track fusion
hierarchial fusion
Kalman filtering
RHV mode
jerk mode
dimension extention algorithm
