带相关噪声的观测融合稳态Kalman滤波算法及其最优性  

Measurement Fusion Steady-State Kalman Filtering Algorithm with Correlated Noises and Its Optimdity

作　　者：顾磊[1] 惠玉松[1] 邓自立[1] 

机构地区：[1]黑龙江大学自动化系,哈尔滨150080

出　　处：《科学技术与工程》2008年第2期328-332,共5页Science Technology and Engineering

基　　金：国家自然科学基金(60374026)资助

摘　　要：对于带相关的输入白噪声和观测白噪声及相关观测白噪声的多传感器线性离散定常随机系统,用加权最小二乘(WLS)法提出了一种加权观测融合稳态Kalman滤波算法,并基于信息滤波器证明了它同集中式观测融合稳态Kalman滤波算法功能的等价性。因而,它具有渐近全局最优性,且可减少计算负担。一个跟踪系统数值仿真例子验证了它的功能等价性。For the muhisensor linear discrete time-invariant stochastic control systems with correlated input and measurement white noises, and with correlated measurement noises, a weighted measurement fusion steady-state Kalman filtering algorithm is presented by using the weighted least squares （WLS）method. Based on the information filter, it is proved that it is functionally equivalent to the centralized measurement fusion steady-state Kalman filte- ring algorithm, so that it has asymptotic global optimality, and can reduce the computational burden. A numerical simulation examples for a tracking systems verifies its functional equivalence.

关 键 词：多传感器信息融合 加权观测融合 相关噪声 稳态Kalman滤波 渐近全局最优性 

分 类 号：O211.64[理学—概率论与数理统计]