Hybridized Spacecraft Attitude Control via Reinforcement Learning using Control Moment Gyroscope Arrays

Abstract

Machine learning techniques in the form of reinforcement learning (RL) can solve complex nonlinear problems found within spacecraft attitude determination and control systems (ADCS). Three CMG arrays were implemented in two simulated spacecraft environments using a reinforcement learning controller. The performance of the controllers were evaluated using target profiles from traditional control law implementations, singularity measure, and variable initial state values. The current research demonstrates that while RL techniques can be implemented, further exploration is needed to investigate the operational efficacy of an approach for producing comparable performance attributes with respect to traditional control laws.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 25, 2021
Accession Number
AD1139410

Entities

People

  • Cecily C Agu

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • Space

DTIC Thesaurus Topics

  • Air Force
  • Angular Momentum
  • Artificial Intelligence
  • Artificial Intelligence Software
  • Attitude Control Systems
  • Bayesian Networks
  • Computational Science
  • Computers
  • Control Systems
  • Control Systems Engineering
  • Data Science
  • Information Science
  • Machine Learning
  • Neural Networks
  • Reinforcement Learning
  • Spacecraft
  • Supervised Machine Learning

Readers

  • Neural Network Machine Learning.
  • Robotics and Automation.

Technology Areas

  • AI & ML
  • AI & ML - Autonomous Systems
  • AI & ML - Bayesian Inference
  • Space
  • Space - Spacecraft Maneuvers