Optimal and Robust Neural Network Controllers for Proximal Spacecraft Maneuvers

Abstract

In this research, reinforcement learning techniques are combined with traditional direct shooting methods to generate optimal proximal spacecraft maneuvers. Open- and closed-loop controllers, parameterized by neural networks, are developed for terminally constrained, fuel-optimal relative motion trajectories using three different thrust models. Neurocontroller performance robustness to parametric uncertainty and bounded initial conditions is assessed. This research demonstrates that neurocontrollers offer a flexible and robust alternative approach to the solution of complex controls problems in the space domain and present a promising path forward to more capable, autonomous spacecraft.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 21, 2019
Accession Number
AD1073578

Entities

People

  • Brandon C. George

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • Space

DTIC Thesaurus Topics

  • Air Force
  • Algorithms
  • Artificial Intelligence Software
  • Artificial Satellites
  • Computational Science
  • Computer Languages
  • Computer Programming
  • Equations Of Motion
  • Genetic Algorithms
  • Geosynchronous Orbits
  • Heuristic Methods
  • Machine Learning
  • Neural Networks
  • Optimization
  • Orbital Elements
  • Spacecraft
  • United States Government

Readers

  • Neural Network Machine Learning.
  • Operations Research
  • Robotics and Automation.

Technology Areas

  • AI & ML
  • AI & ML - Autonomous Systems
  • AI & ML - Machine Learning Algorithms
  • AI & ML - Neural Networks
  • Space
  • Space - Spacecraft Maneuvers