Nonlinear Transfer Alignment Model and Filtering Algorithm for Strapdown Inertial Navigation System of Shipborne Weapons

To satisfy the requirement of fast reaction and high precision of initial alignment for shipborne weapons, and to enhance the transfer alignment for shipborne weapons with large misalignment, we present a method of combining the rapid transfer alignment model based on quaternion and the nonlinear unscented Kalman filter (UKF) algorithm, and establish a nonlinear error model for strapdown inertial navigation system(SINS) of shipborne weapons. This error model introduces quaternion to describe the attitude errors in order to enhance the speed and accuracy in calculating attitude. We select a velocity and attitude matching approach to improve the observability of the system, and solve the ill-condition of covariance matrix of the traditional nonlinear UKF filter by using singular value decomposition(SVD) to ensure the algorithm robustness. The simulation results show that the proposed method can effectively improve the precision and speed of transfer alignment for shipborne weapons with large misalignment, and its calculation accuracy and alignment time meet the design requirements of the system.