摘要
针对Kalman滤波器在捷联惯导系统(SINS)初始对准中的应用,系统分析了Kalman滤波器参数(包括估计误差协方差阵初值P0,模型噪声方差阵Q和量测噪声方差阵R)选取对系统状态变量的估计精度和收敛速度的影响。采用协方差性能分析法,进行了Kalman滤波器参数优化仿真,仿真结果表明:调整P0的取值可改变状态变量估计的收敛速度,调整Q或R的取值,既可改变状态变量(尤其是陀螺误差)的收敛速度又可改变它们的估计精度。综合考虑时,P0的取值要比真实值大一些,Q和R的取值要比真实值小一些,这样既可缩短陀螺误差和加速度计偏置误差的估计时间,又可提高它们的估计精度。文中还给出了使滤波器正常可靠工作的P0、Q和R参数的范围。
This paper analyzes the effects of the choice of the Kalman filter parameters（including the initial values of estimation error variance P0, the values of measurement noise variance R and the model noise variance Q） on state variables estimation performance of SINS during initial alignment. The optimization of parameters is simulated based on covariance performance analysis approach. The results show that a different P0 matrix yields a different magnitude transient characetristic of state variables （especially gyro drift errors） estimation, and changing Q or R affects the transient duration and the steady state operation of gyro drift errors and the accelemeter bias errors estimation. Increasing P0, decreasing Q and R can not only shorten the errors estimation time, but also aquire good estimation accuracy. At last, the paper presents the parameter ranges of P0 ,Q and R which keep the Kalman filter working normally and reliablly.
出处
《中国惯性技术学报》
EI
CSCD
2006年第4期12-17,共6页
Journal of Chinese Inertial Technology
关键词
捷联惯导
初始对准
速度匹配
参数优化
滤波调谐
SINS
initial alignment
velocity matching
parameters optimization
filter tuning
