Orbit Prediction

Orbit prediction aims to accurately forecast the future positions and velocities of space objects, crucial for collision avoidance and space situational awareness. Current research heavily emphasizes using machine learning, particularly deep learning and Gaussian processes, alongside traditional methods like Kalman filters, to improve prediction accuracy and efficiency, especially when dealing with non-conservative forces like atmospheric drag and the complexities of closely-spaced objects. These advancements are driven by the increasing number of space objects and the need for more precise and computationally efficient orbit determination, impacting space debris mitigation, satellite navigation, and the overall safety of space operations.