Effect of Adaptive Fading Scheme on Invariant EKF for Initial Alignment Under Large Attitude Error and Wave Disturbance Condition

This paper proposes an invariant extended Kalman filter (IEKF) to which an adaptive fading scheme is applied as an initial alignment technique for marine vehicles in mooring condition. In the presence of wave disturbance, the coarse alignment accuracy decreases, which affects the initial attitude estimation error of the fine alignment. Accordingly, the IEKF robust to the large initial errors is applied during the fine alignment process. Moreover, the accuracy of zero velocity updates (ZUPT) decreases because they have non-zero translational velocity under the influence of waves. Thus, the performance of the conventional EKF using a fixed measurement covariance is degraded. In this paper, the adaptive fading scheme adjusting Kalman gain is applied to the IEKF to accelerate the convergence rate. The proposed method is verified by numerical simulation.