摘要
通过大量的仿真，发现现行的滤波算法各有其利弊．在综合考虑和分析各种滤波方法之后，在理论推导的基础上，提出了一种新的控制滤波发散的方法，即通过判定发散的判据，求出噪声统计的估计值，然后再按得到的噪声统计的估计值计算新息序列的协方差阵．将该算法应用于惯导初始对准问题．通过仿真，可以验证该算法是十分有效的．
By a lot of simulation, some insufficiencies of filtering methods now existing were found.After these methods were analyzed, a new filtering design method is proposed based on theoretical deduction. According to the criterion for controlling .divergence. a estimate of statistic of measurement noise can be obtained, and the covariance of innovation sequence is calculated using the estimated statistic of measurement noise. The method is applied to the design of the initial alignment of a inertial navigation system, and the simulation results prove that the new design for controlling divergence is useful.
出处
《北京航空航天大学学报》
EI
CAS
CSCD
北大核心
1994年第4期392-396,共5页
Journal of Beijing University of Aeronautics and Astronautics
关键词
卡尔曼滤波
发射
准直
惯性导航
航空
Kalman filtering
divergence
alignment
methods
criterion
