Optimal control co-design of attitude control systems

Attitude control is the process of orienting a spacecraft toward a desired point in space. Conventional systems for attitude control have been designed without considering the compliant response of the spacecraft. This leads to complex control algorithms that avoid exciting vibrational modes of the spacecraft structures, particularly deployable panels. Most conventional approaches try to modify the deployable structures to reduce mechanical compliance, which comes at a mass and cost penalty. The performance of such systems can be greatly improved if the system is optimized from both a design and control perspective simultaneously, known as control co-design (CCD). This dissertation proposes new mathematical modelling methods and CCD optimization strategies to support the creation of novel Attitude Control System (ACS) that utilizes mechanical compliance of deployable panels to provide attitude control. New contributions to optimization methods utilize reinforcement learning algorithms adapted for CCD problems. These algorithms can aid in quickly exploring the design space of novel systems and provide design rules to help design systems which are close to optimal without needing to solve a new optimization problem each time. These new methods will be used to develop Strain Actuated Solar Arrays (SASA), an ACS that offers attitude stabilization capability at the nano-radian scale by actively canceling mechanical vibrations on the spacecraft. The precise levels of pointing are enabled by designing the system while considering the control trajectories that SASA is expected to execute. The co-design studies will then be used to expand the capabilities of the SASA system to not only provide attitude control at the small angles, but also to provide attitude maneuvering capability for arbitrarily large angles. This extension to SASA is called Multifunctional Structures for Attitude Control (MSAC), which works like a momentum exchange device and can serve as a replacement to conventional reaction wheel assemblies and control momentum gyroscopes. The dissertation concludes with hardware-in-the-loop validation of SASA and MSAC prototypes designed using the new models and methods. The response of these prototypes is compared against the predicted performance and serves as a validation for the new models and techniques developed.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Aerospace Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2021

Author dc:creator

• Vedant

Contributors dc:contributor

• Allison, James T
• Ghosh, Alexander M
• Woollands, Robyn
• West, Matthew
• Waldrop, Lara

• Attitude control
• Control co-design
• Optimization
• Reinforcement learning

Statement dc:rights

• Copyright 2021 Vedant

Language dc:language

en

Handle dc:identifier

https://hdl.handle.net/2142/117875

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/117875