基于递推最小二乘的航姿系统罗差校正

Magnetic deviation compensation using recursive least square for AHRS

为校正姿态与航向参考系统使用环境中软硬铁产生的罗差,提出了一种利用地磁场连续转动信息补偿三轴软硬铁的方法。由于陀螺短时精度高,初始对准后短时内可得到精度较高的捷联矩阵,如果已知当地磁倾角,由捷联矩阵可将已知的地磁场投影到载体系。当航姿系统改变姿态角连续转动时,采用递推最小二乘算法来估计12个罗差参数。实验说明,与常用罗差校正方法相比,此方法在全姿态范围内效果更好,水平时航向角精度为0.3°,35°倾角航向角精度为1.1°,59°倾角航向角精度为1.2°。

To compensate the magnetic deviation produced by the soft iron and hard iron distortion in attitude and heading reference system（AHRS）, a recursive least square （RLS） algorithm using continuous magnet field information was developed. The strapdown matrix can achieve high precision for a short-term after AHRS＇s initial alignment. With the local inclination angle of magnetic field being known, when AHRS changes attitude and rotates one cycle, the magnetic field vector can be projected onto the body frame. The 12 magnetic deviation parameters, which are equivalent in influence to that of soft iron and hard iron distortion, can be estimated by RLS. The experiment results demonstrate that the performance of RLS is better than that of common magnetic deviation compensation methods in the whole attitude range with the heading accuracy being 0.3°, 1.1°, and 1.2° for tilt angle of 0°, 35°, and 59°respectively after compensation.