Gyroscope drift estimation analysis in land navigation systems

The objective of this paper is to analyze the estimation performance of gyroscope drifts in integrated land navigation systems, while system is stationary. Inertial navigation system error dynamic is augmented by inertial sensor bias and drifts to improve the integration performance. To estimate the augmented error model states GPS position is used. To assess estimation performance Kalman filter covariance analysis is used. While aiding INS with GPS in different motion scenarios of land vehicles, it is observed that down gyro drift estimation is very weak. When land vehicle is stationary, a zero turning update algorithm (ZTUPT) is proposed to improve observability of the down gyro drift. Down gyro drift estimation is simulated by using ZTUPT algorithm. It is seen that while calibrating down gyro drift, ZTUPT works better than position update.