Autonomous Distributed Angles-Only Orbit Determination using Multiple Observers
Abstract
The primary goal of this proposal is the derivation of new dynamics models and estimation algorithms to enable autonomous real-time detection, tracking and orbit determination of many resident space objects using bearing angle measurements obtained by multiple cooperative observer satellites. As such, this proposal responds to the need of the Department of Defense to autonomously detect and perform accurate orbit determination for large numbers of uncooperative resident space objects. This capability would enable autonomous navigation in GNSS-denied environments, improved data products and association for space situational awareness, and new classes of distributed space missions. The addressed problem is challenging because targets must be tracked across multiple images from sensors at different locations, and bearing angle measurements provide no range information, resulting in weak observability. We propose to overcome these challenges through new algorithms for target identification and tracking that exploit domain-specific knowledge of perturbed orbital motion to provide accurate measurement assignment without a-priori state information. A new quantitative nonlinear observability analysis is proposed to determine conditions under which the orbits and auxiliary parameters of observers and targets are observable. The findings of this analysis will inform the formulation of a new batch estimation algorithm that provides initial state estimates for detected targets with less than one orbit of sparse measurements. Finally, a new adaptive uncertainty-aware sequential filter capable of detecting and estimating unknown orbit maneuvers will be developed and validated in hardware-in-the-loop simulations. These algorithms rely on new dynamics models for Earth and cislunar orbits that leverage a representation of orbital motion in integration constants to minimize computation cost and maximize accuracy. Overall, the proposed methods will enable autonomous, self-contained absolute orbit determination of uncooperative targets using simple orbiting sensors in a distributed manner, providing the United States with new strategic capabilities for surveillance, space situational awareness, and proximity operations.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 21, 2022
- Source ID
- FA95502110414XX0
Entities
People
- Simone D'Amico
Organizations
- Air Force Office of Scientific Research
- Stanford University
- United States Air Force