带相关噪声的加权观测融合估计算法及其全局最优性  

Weighted measurement fusion estimation algorithm with correlated noises and its global optimality

作　　者：王欣 朱齐丹[1] 孙书利[2] 

机构地区：[1]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001 [2]黑龙江大学电子工程学院,黑龙江哈尔滨150080

出　　处：《系统工程与电子技术》2010年第10期2057-2061,共5页Systems Engineering and Electronics

基　　金：国家自然科学基金(60874062);教育部科学技术重点项目(209038);黑龙江省教育厅重点实验室项目(DZZD20100038)资助课题

摘　　要：针对多传感器线性离散定常随机控制系统,当具有相关噪声且每个传感器带不同观测阵时,基于矩阵满秩分解与加权最小二乘理论,提出了新的加权观测融合估计算法。该算法首先将多个传感器的观测折算到一个等效的传感器上,对等效的传感器系统进行估计,证明了其估计结果相同于集中式融合稳态Kalman估计结果,因而它同样具有渐近全局最优性,且可明显减小计算负担,便于实时应用。仿真实验结果表明了该算法的有效性。In view of the multisensor linear discrete time-invariant stochastic control system with correlated noises and different measurement matrices for every sensor,a new weighted measurement fusion estimation algorithm is presented by using the full-rank decomposition of matrix and the weighted least square theory.The newly presented algorithm firstly converts the measurements of many sensors into an equivalent sensor,which is then estimated.The estimating result is proved to be equivalent to the centralized fusion steady-state Kalman estimating result,so that it also has the asymptotic global optimality.It can obviously reduce the computational burden,so it is convenient for application in real time.A simulation result shows the effectiveness of the proposed algorithm.

关 键 词：相关噪声 满秩分解 加权观测融合 集中式融合 

分 类 号：O211.64[理学—概率论与数理统计]