摘要
针对采用卡尔曼滤波算法进行线性系统状态估计的应用场合,提出了可观测度实时估计方法。该方法通过提取离散卡尔曼滤波迭代计算过程中的矩阵进行分析,因而在无需增加额外计算量的情况下,可获得系统可观测度性能的实时估计。该方法可以指导滤波器的设计过程或者在线估计系统的滤波器性能,具有工程应用价值。给出了应用该方法的原理推导过程,通过SINS系统地面单/双位置对准的实例比较分析证明该方法的有效性。
In applications of state estimation in a linear system via kalman filter, a real time estimation method of degree of observability was proposed. Degree of observability could be obtained through matrix analysis during iterative computation of discrete Kalman filter without extra computation. The proposed method is very useful for filter design or online performance estimation. Principle of this estimation method was derived and efficiency was proved by contrast between single and double position ground alignment process of strapdown inertial system.
出处
《宇航学报》
EI
CAS
CSCD
北大核心
2009年第3期982-987,共6页
Journal of Astronautics
关键词
线性系统
可观测度
卡尔曼滤波
组合导航
两位置对准
Linear system
Degree of observability
Kalman filter
Integrated navigation
Double position ground alignment
