摘要
提出了一种基于对角阵加权的改进联邦滤波算法。根据该方法,在全局融合过程中只需计算加权对角阵的每个标量元素,避免了加权矩阵的求逆运算,提高了计算稳定性,减少了计算量。因此,非常有利于实时应用。仿真算例说明该方法是有效的。
An improved federated filtering algorithm is given based on diagonal matrix weighting. By this method, only computing each scalar component of the weighting diagonal matrixes in the global fusion process is needed. Hence, computing the inverse of the covariance matrixes is avoided, which improves the numerical stability of filtering process and reduces the computation load. It is highly advantageous to realtime applications. Simulation shows the effectiveness of this method.
出处
《传感技术学报》
EI
CAS
CSCD
北大核心
2006年第2期497-500,共4页
Chinese Journal of Sensors and Actuators
基金
空军工程大学博士论文创新基金资助
关键词
对角阵加权
联邦滤波
信息融合算法
实时应用
diagonal matrix weighting
federated filter
information fusion algorithm
real-time application
