Is this project an undergraduate, graduate, or faculty project?
Graduate
group
What campus are you from?
Daytona Beach
Authors' Class Standing
Chinmay Mirji, Graduate Student Saeed Ahmadi, Graduate Student
Lead Presenter's Name
Chinmay Mirji
Faculty Mentor Name
Hao Peng
Abstract
Conventional attitude control algorithms often degrade when faced with actuator faults, sensor noise, or system uncertainties. This work presents a reinforcement-learning (RL) framework for satellite attitude recovery under unknown failures, focusing on real-time deployment through a processor-in-the-loop (PIL) setup. A continuous-control DDPG agent is trained in a high-fidelity Python/Basilisk simulation environment, where domain randomization captures variations in inertia, external torque, and actuator limitations to promote robust policy learning.
Did this research project receive funding support from the Office of Undergraduate Research.
No
Reinforcement Learning - Driven Satellite Attitude Recovery: Unknown Faults, Simulation-to-Processor in Loop
Conventional attitude control algorithms often degrade when faced with actuator faults, sensor noise, or system uncertainties. This work presents a reinforcement-learning (RL) framework for satellite attitude recovery under unknown failures, focusing on real-time deployment through a processor-in-the-loop (PIL) setup. A continuous-control DDPG agent is trained in a high-fidelity Python/Basilisk simulation environment, where domain randomization captures variations in inertia, external torque, and actuator limitations to promote robust policy learning.