一种五轴冗余惯导的系统级标定方法研究

A system-level calibration method for five-axis redundant inertial navigation system

关键词:  系统级标定  Kalman滤波  冗余惯性测量单元  零空间约束

Key words:  System-level calibration  Kalman filter  Redundant inertial measurement unit (RIMU)  Null space constraint

摘要:

针对五轴冗余惯性测量单元（RIMU）的系统级标定问题，推导了简化线性标定误差模型。载体坐标系下的运动参数矢量通过转换关系投影至传感器坐标系,并由RIMU敏感得到。当存在安装误差时，将RIMU的输出通过转换矩阵转换至载体坐标系，其安装误差会在载体坐标系中得以体现。因此，五轴冗余器件的安装误差补偿问题转换为三轴正交配置下的器件安装误差补偿问题。同时，加入了零空间约束以提高标定系统状态量的可观测度，采用１９位置转动方案，并通过Kalman滤波进行估计。该标定方案可一次性准确标定出RIMU中陀螺仪和加速度计的所有标定参数，通过仿真实验验证了其可行性。

Abstract:

To address the system-level calibration problem of a five-axis redundant inertial measurement unit (RIMU), a simplified linear calibration error model is derived. The motion parameter vector in the body frame is projected onto the sensor frame via the transformation relationship and sensitized by the RIMU. In case of an installation error, the output of the RIMU is converted to the body frame by the transformation matrix, and the installation error is reflected in the body frame. Thus the installation error compensation problem of five-axis redundant devices is transformed into the installation error compensation problem of devices in the three-axis orthogonal configuration. In addition, zero-space constraints are incorporated to improve the observability of the state quantity of the calibrated system. A 19-position rotation scheme is employed and the estimation is performed by Kalman filter. Simulation results show that this calibration approach can accurately calibrate all gyroscope and accelerometer parameters in RIMU simultaneously.