Advancements in Space Object State Estimation During Constant Thrust Maneuver Using Distributed Sensor Networks: A Radar Based Approach

The increasing number of Resident Space Objects (RSOs), encompassing operational satellites and space debris, presents a significant challenge to space mission safety and efficiency. Operational satellites undergo numerous maneuvers throughout their lifetimes, necessitating the estimation of new orbits for cataloging purposes. This process presents a substantial challenge in terms of maintenance. Traditional approaches to maneuver estimation depend on observing post-thrust alterations in the orbit’s semi-major axis, a method lacking real-time capability. Although ground-based sensors like radar provide essential real-time tracking data, their deployment incurs significant costs. Therefore, the feasibility of a cost-effective solution is explored by using combinations of already existing assets, such as radar and radio telescopes, to enhance the real-time accuracy of satellite tracking. In this study, the trilateration technique, traditionally used for Initial Orbit Determination (IOD) in the case of multiple sensors, is modified by introducing a fresh set of equations for the initial estimation of thrust maneuver. Additionally, it is shown how this refined method plays a crucial role in facilitating more accurate and reliable convergence within the Unscented Kalman Filter (UKF) framework. Simulation of the International Space Station orbit with constant tangential thrust has been undertaken, followed by Root Mean Square Error analysis for performance evaluation. Furthermore, the absolute percentage error in Keplerian parameters is calculated to validate the methodology, revealing that all parameters exhibit an error of less than 1%. The results show significant IOD and tracking accuracy improvements, offering a promising direction for future Space Surveillance and Tracking (SST) efforts.