A Novel Method for Fast Stationary Initial Alignment Based on Extended Measurement Information

• Main Authors: Haotian Yang, Bin Zhou, Lixin Wang, Qi Wei, Rong Zhang
• Format: Article
• Language: English
• Published: IEEE 2019-01-01
• Series: IEEE Access
• Subjects: Initial alignment; Kalman filters; measurement equation; observability; state estimation
• Online Access: https://ieeexplore.ieee.org/document/8897663/

This paper presents an innovative complete observable initial alignment method for strapdown inertial navigation system. This method negates the need for multi-position rotation or a complex rotating mechanism. First, the coupling relationship between the error state variables is analytically derived. Based on the equivalence relationship between the acceleration output and the angular rate output and error state variables, an improved extended-measurement equation is then established. The feasibility of the scheme that accelerates the convergence speed is theoretically demonstrated, and the observability of the proposed method is qualitatively and quantitatively analyzed in comparisons with established methods (piece-wise constant system, singular value decomposition, and error covariance matrix analysis). The results confirm that the improved measurement model enhances the observability of state variables to different extent and can achieve complete observability of the system. The proposed method not only can provide the fast and accurate estimation of the alignment error, but also can predict the gyro bias online. More specifically, the equivalent horizontal accelerometer output certainly accelerated the convergence of the horizontal alignment error, and introducing the equivalent gyroscope output improved the convergence speed of the heading alignment error. It is a potential candidate for speeding up the convergence of the stationary initial alignment.