Performance Comparison of Two Filtering Approaches for INS Rapid Transfer Alignment

This paper presents a velocity matching scheme for rapid transfer alignment of INS. Velocity matching algorithm is known as one of the best matching algorithm of SINS in-motion alignment since it can achieve high alignment accuracy without lengthy time requirement. We try to find a more excellent filter than the traditional kalman filter in this paper. Firstly the mathematical model of velocity matching is presented, and the kalman filter and robust filter formulations are introduced. Then it analyses the observability of various states in velocity matching scheme. This paper mainly focuses on the misalignment estimated accuracy and convergence speed of the proposed algorithm. In order to validate the effect of the velocity matching algorithm, three-axis sway model simulation based on 10 dimensions states and 5 dimensions states is carried out. The results of simulation suggest that the estimating accuracy of kalman filter is equivalent with that of robust filter, but convergence rate of robust filter are much better than that of kalman filter. Furthermore, reduced order model is adaptable for the practical INS alignment considering the computational complexity of the presented velocity matching algorithm.