Hardware-in-the-Loop Verification of Model Predictive Control of Rotation Floating Space Robot for Fine Manipulation

摘要

Autonomous robotic systems are expected to play a very significant role in future space sustainability missions like on-orbit servicing, active debris removal, assembly, etc. Rotation Floating space robots are a newer class of space robots wherein the position and orientation of a robotic arm mounted on-board a floating satellite is actively controlled along with simultaneous control of the satellite base orientation. Modeling such systems with a high degree of non-linearity and coupled degree of freedom (DOF) is difficult. We have, in our previous works, modeled a Free Floating space robot in a simulation environment in PyBullet and used a Model Predictive Controller for position and orientation control of a UR5 robotic arm mounted on the base satellite. Such works are however missing for Rotation Floating Space Robots. Also, experimental validation of such systems is difficult due to the difficulty in replicating zero gravity environment and providing 6 DOF motion to robots on the ground. This work thus addresses these two issues. We model a Rotation Floating space robot in PyBullet and use an MPC-based layered control architecture for simultaneous 9 DOF control. The simulation shows the complete workflow from the time the robotic arm is deployed till the target is successfully captured or docked on. Thereafter, we adopt a hybrid approach for Hardware-in-the-loop validation of the simulations performed in PyBullet. One robot replicates the chaser arm motion while one robot is used to replicate the 6 DOF motion of the target body. Both the results are presented in the paper.