多传感器线性最小方差最优信息融合估计准则

Multi-sensor Optimal Information Fusion Criterion in Linear Minimum Variance Sense

用Lagrange乘数法和矩阵微分运算 ,分别提出了按矩阵加权、按标量加权和各分量按标量加权的三种线性最小方差信息融合估计准则 ,其中考虑了估计误差之间的相关性 ,推广和发展了现有文献的结果。文中比较了三种融合估计的精度和计算负担 。

Using Lagrange multiplier method and matrix differential operation, three information fusion estimation criterions are presented in the linear minimum variance sense, where the fusion estimators are respectively weighted by matrices, weighted by scalar and weighted by scalar on components. The correlation among estimation errors is considered. The results of existing literatures are extended and developed. Their precision and computational burdens are compared. They can be applied into optimal information fusion estimation for the states or signals.