Orbit Determination for Continuously Maneuvering Starlink Satellites Based on an Unscented Batch Filtering Method

Area of Science:

• Aerospace Engineering
• Astrodynamics
• Orbital Mechanics

Background:

• Orbit determination for non-cooperative low Earth orbit (LEO) objects with continuous low-thrust maneuvers is complex.
• Large satellite constellations like Starlink present unique challenges for tracking and orbit determination.

Purpose of the Study:

• To develop and validate a robust method for orbit determination of maneuvering LEO objects.
• To enhance space situational awareness for satellite constellations by providing accurate orbit knowledge.

Main Methods:

• Integration of unscented transformation into a batch filtering framework.
• Utilizing an optimized rho-minimum sigma points sampling strategy.
• Employing a reduced dynamics model with Earth's non-spherical gravity and equivalent along-track acceleration for low-thrust and atmospheric drag.

Main Results:

• The proposed method demonstrates reliable convergence within 10 iterations.
• Achieves 24-hour position prediction root mean square errors below 3 km under realistic noise.
• Validation with a higher-fidelity model confirms improved accuracy and robustness.

Conclusions:

• The developed method provides accurate orbit determination for continuously maneuvering LEO satellites.
• It offers a valuable tool for space situational awareness, particularly for constellations like Starlink.
• The approach is effective even without prior knowledge of along-track acceleration.