Attitude Estimation Method of High-Spinning Body Through Onboard Sensor Fusion Under Uniaxial Gyroscope Saturation Environment

Abstract

This article deals with a 3-D attitude estimation problem of a dynamic object that involves a uniaxial, high-speed rotational motion. As a high spinning dynamics typically causes an angular rate saturation in most gyros, it is also challenging to determine the object's attitude of the principal rotation axis and correspondingly the attitude of other orthogonal axes. To solve this problem, this article proposes two techniques to accomplish the attitude estimation under uniaxial gyro saturation condition. First, this article develops a new integrated filter scheme based on a reduced state model and attitude kinematics. The proposed estimation algorithm takes advantage of the magnetic measurement and gyro's partial rates to obtain a posteriori attitude estimate in the orthogonal axes and further develops to derive the principal axis attitude deterministically. Despite decent estimation performance, a measurement degeneracy may be caused in the reduced state model theoretically. To suppress the degeneracy effect, this study also exploits an augmented attitude estimation method, which adopts a post-processing computational rate in the principal axis through a time-frequency analysis on magnetic measurements. Finally, the proposed technique's performance was demonstrated through both simulation and experiment with the high spinning rate of 10846 and 2880°/s, respectively, where the error level was measured at the sub-degree in the orthogonal axes and slightly higher in the principal axis.