基于平滑算法的惯导系统误差校正技术

惯性导航系统受初始误差的影响会产生周期性振荡以及发散误差。针对这一问题,提出了基于自适应衰减卡尔曼滤波及固定点平滑的估计算法,对系统初始误差进行估计,估计结果代入系统误差模型推导误差发散情况,以此对系统输出进行补偿。仿真结果表明，该算法可以有效抑制惯性导航系统输出误差的振荡及发散。

Error calibration technology of inertial navigation system based on smoothing algorithm

Affected by initial error, the inertial navigation system (INS) will generate periodic oscillation and divergence error. To deal with this problem, an estimation algorithm based on adaptive fading Kalman filter (AFKF) and fixed point smoothing (FPS) is proposed. The initial error is estimated, and the divergence error is derived from the error model of INS, in order to calibrate the output of INS. Simulation results show that the proposed algorithm can restrain the oscillation and divergence of INS errors effectively.