2025-09-09 –, Audi-Midi
This work describes a novel hybrid reinforcement learning enhanced trajectory planning algorithm for an active debris removal scenario for combined control of a satellite with a 7-axis robotic arm. A reinforcement learning algorithm is combined with a correction algorithm and classical trajectory planning to handle the collision free approach of a chaser satellite to a target, and placing the gripper at the robots near the grasping point for use with a combined controller, which commands the satellite and its robotic arm simultaneously. The algorithm is verified using a complex simulation scenario study implemented in Modelica/FMI.
