SINS In-Motion Initial Alignment Based on Backtracking Multi-vector Attitude Determination

Backtracking alignment is essentially a data multiplexing technology, which uses reverse navigation and data fusion to obtain better results than traditional methods. Since the multi-vector attitude determination (MAD) method is widely used in the in-motion alignment, it can be considered to combine this method with backtracking to further improve alignment accuracy. In this paper, an in-motion alignment method based on backtracking multi-vector attitude determination (BMAD) is proposed. Part of results in the forward process are stored and reused in the reverse alignment. Then an information fusion method based on iterative matrix eigenvalues is further proposed to improve yaw alignment accuracy. Simulation experiment is designed and compared with backtracking alignment based on Kalman filter to verify the effectiveness of the method.