ABSTRACT
Transfer alignment is usually applied to weapon systems that need a rapid launch and complex external environment. In actual conditions, the misalignment angles are usually not small and the error model cannot be processed simply by linear filtering for large misalignment angles. In this paper, a large misalignment angle error model based on the multiplicative quaternion is derived, and a quaternion strong-tracking fifth-order cubature Kalman filter is proposed. To keep the quaternion’s norm unchanged during the filtering process, a quadratic Padé approximation method is used in the quaternion iterative decoding process. The adaptive fading factor from a strong tracking filter is added to the time update and measurement update equations from the fifth-order quaternion cubature Kalman filter; thus, the gain matrix can be adaptively adjusted, and the filtering divergence can be effectively prevented. Simulation results show that the proposed algorithm can solve the problem of transfer alignment for large misalignment angles, and the proposed algorithm can improve the transfer alignment accuracy and rate of convergence in large misalignment angle cases.
