A transfer alignment method for inertial attitude correction of spherical motor on moving base

Inertial attitude measurement of spherical motor on moving base (SMMB) confronts the problems of unobservable azimuth and acceleration interference. This paper presents a transfer alignment method for inertial attitude correction of SMMB with specific exterior excitation. Because of the low friction torque of the spherical joint in the SMMB, the rotation is not able to be wholly transferred from base to rotor, which is required in the common method of velocity-attitude matching. With an explicit model about the acceleration transfer, the accumulation of unobservable error based on nominal-state is derived. Then the transfer alignment algorithm with innovation-based adaptive filter is proposed. The method brings external acceleration measurement into regular attitude estimation and adopts both estimation of unobservable error and acceleration transfer error into adaptive filter for the balance of azimuth correction and tilt error suppression. The simulation and analysis show that the method can correct the unobservable azimuth angle in high amplitude and frequency excitation as well as better tilt error suppression for transfer error with 20% uncertainty. The method is verified in a prototype platform.