Steering logic of control moment gyros using artificial potential function approach

Abstract

A Control Moment Gyro (CMG) cluster for an agile spacecraft usually suffers from a singularity problem where it does not meet a desired torque. To solve the problem, a proposed steering law consists of SR (singularity robustness) inverse and SR null motion to compensate each other. The SR null motion solves the singularity problem that the SR inverse stays in the singularities. Especially, various potential functions are included in this dissertation, which newly consider the distance between the current gimbal angle and the gimbal angle corresponding to all singularities. All the combinations of singular gimbal angles are regarded as a target and obstacles of the potential function. After a spacecraft maneuver, an initial gimbal angle is regarded as the target to secure the repeatability of next maneuver. When avoiding or escaping the obstacles corresponding to the unwanted singular gimbal angles, the proposed steering law rapidly changes the gimbal angle by taking torque error. To overcome the local minima problem which is generally a serious drawback in using the potential function, an algorithm for changing parameters of the potential function are proposed in the steering law. To avoid the internal singularity, moreover, the present study introduces a configuration change of the CMG cluster. The proposed steering law is mathematically analyzed and verified through numerical simulations. It shows that the CMG cluster can avoid an internal singularity rapidly, resulting in faster maneuver time of the spacecraft.