Application of UKF/KF in Single-axial Rotation Initial alignment

Abstract

In order to improve the initial alignment precision on static base is restricted because the system is not completely observable, UKF used in large misalignment angle has more calculate amount, the scheme of single-axial rotation is put forward, and UKF/KF feedback algorithm is applied to reduce calculate amount. The continuous rotation nonlinear error model is deduced out when the azimuth misalignment angle is large, attitude matrix is updated by equivalent rotation vector attitude algorithm, and UKF/KF is used to complete the process of precise alignment. UKF is used to state estimate when model is nonlinear error model, when misalignment angle is reduced less than one static value by feedback, error model can be considered as linear model and KF is used to complete state estimate. The simulation result shows that the alignment estimation accuracy in single-axial rotation scheme is obviously higher than two-position alignment and static alignment, UKF/KF has the same alignment precision and speed compare with UKF, but the speed of compute is enhanced.