In-Field Calibration Method for DTG IMU Including g-Sensitivity Biases

Navigation performance can be significantly enhanced by the inertial measurement unit (IMU) calibration. Aiming at the dynamically tuned gyro IMU, a novel in-field calibration method is proposed in this paper. Taking gravity and earth's rotation rate as a reference, all accelerometer errors, the constant biases, and the g-sensitivity biases of gyro can be calibrated using the unscented Kalman filter based on the norm-observed method. The scale factor errors and installation errors of gyro can be estimated though the Kalman filter with velocity errors as observation. In order to guarantee that all error parameters are observable, a 26-position and its optimized 12-position rotation schemes are designed. The proposed method can achieve correct calibration without a high-precision turntable. Besides, it can improve the ability of analysis and the deviation of interference sources when some error parameters are calibrated unreasonable. The simulations and in-field calibration experiments are conducted to verify the effectiveness of the proposed method. The simulations demonstrate that the calibration accuracy of two schemes is almost equivalent. Compared with the system-level calibration, the calibration results of two schemes are within a reasonable range. The semi-physical simulations indicate that finding out and separating interference sources can be achieved easily with the proposed method. Thus, the proposed method has great application values in in-field calibration.