A rapid and high-precision initial alignment scheme for dual-axis rotational inertial navigation system

In the initial alignment process, it is crucial to realize rapid and high-precision estimation of filter states, especially to the misaligned azimuth angle and vertical sensor's errors in long-term application. Benefiting from the gimbals structure in dual-axis rotational inertial navigation system (RINS), the state's observability could be highly improved by maneuvering the two gimbals in a certain order. In this paper, two principles in designing rotation schemes for alignment are proposed. Based on these principles, a novel rotation scheme is introduced integrating the advantages and overcoming the limitations of traditional alignment methods. A Kalman Filter is established and the observability of misaligned azimuth angle and vertical sensor's errors are discussed basing on the piece-wise constant system observability analysis method, singular value decomposition analysis method and covariance matrix analysis method. Simulation and system experiment results verify the correctness of the two principles for designing rotation scheme in alignment and prove the availability of the proposed method in actual dual-axis RINS.