State-Dependent Special Orthogonal Group Filter for SINS In-Motion Initial Alignment

Aiming at the in-motion initial alignment problem of the strapdown inertial navigation system (SINS), this article proposed a precise in-motion alignment method based on special orthogonal group representation and state-dependent filter. Two major innovations are involved in this method. First, the special orthogonal group is implemented to represent the initial attitude matrix, and an equivalent state-dependent in-motion alignment model is developed. The filtering equations are derived under the minimum variance criterion to decrease the influence of the state-dependent noise problem. Second, this method uses the orthogonal projection theorem to deduce the update equation, which retains identity to assure accuracy and stability when eliminating the state-dependent noise. Simulation and experimental results demonstrate that the proposed alignment method has advantages over existing methods in terms of alignment accuracy and alignment time and performs well in the in-motion alignment process of SINS.