Quaternion augmented unscented Kalman filter and its application to rapid transfer alignment under large misalignment
四元数扩维无迹卡尔曼滤波算法及其在大失准角快速传递对准中的应用

An equivalent rapid transfer alignment model based on the multiplicative quaternion is derived for large misalignment measurements on the basis of nonlinear Euler angle error model. In order to apply quaternion to the unscented Kalman filter(UKF), a quaternion augmented UKF(Q-AUKF) algorithm is proposed. The arithmetic problem of multiplicative quaternion noise is solved by augmenting the system noise onto the state vector. An average quaternion method is adopted to guarantee the normalization of weighted quaternion. The Q-AUKF algorithm is applied to the rapid transfer alignment model and the simulation results show that the proposed algorithm has higher filter accuracy and faster convergence speed.