The gravity estimation with square-root unscented Kalman filter in the cold atom gravimeter

In the atomic gravimeter, the interference fringe is obtained by scanning the chirp rate of the Raman lasers to compensate for the gravity induced Doppler frequency shift and then the gravitational acceleration can be obtained by extracting the phase of the atomic interference fringe. The phase of the interference fringe is proportional to the chirp rate, therefore the accurate estimation of the chirp rate is the key issue to obtain accurate gravitational acceleration. In practice, the model for estimating the chirp rate is not perfect, and there is system state error. In this paper, we propose a square root unscented Kalman filter (SR-UKF) approach for estimating gravitational acceleration, which can take the system state error into account. The experiments on simulated and real data are performed to demonstrate the feasibility and effectiveness of the SR-UKF method, which can provide an alternative approach for the gravity estimation. In addition, it can be combined with the tidal model for gravity estimation, and can also be used for real-time gravity measurement.