Title: Statistical methods for space surveillance
Authors: Antonio Arcos - Universidad de Granada (Spain) [presenting]
Abstract: Since the first human spacecraft, Sputnik-1, the Earth proximities have been occupied by multiple human-made objects which have created a particle environment known as ``Space Debris''. The mitigation of this phenomena is a crucial safety task nowadays, therefore adopting measures to relieve and prevent these threats has aroused a worldwide concern. Modern space systems require the highest possible accuracy to work efficiently. Sensors measurements should be differentially corrected to actually determine de real and precise orbit in what is called Statistical Orbit Determination (SOD). The application of predicting filters, as Kalman or particle, allows the system to rapidly and memoryless solve nonlinear least-square problems simultaneously estimating not only the state but also the covariance matrix of the target. Besides the SOD part, other statistical methods are applied to model the different parts of the tracker architecture, solving likelihood optimal associations with several algorithms. A simulated scenario is created to model various space environments and its measurements to conclude in an optimal configuration of the tracker; in which it could be remarked the performance of the Unscented Kalman Filter (UKF) and the Joint Probabilistic Data Association (JPDA) algorithm.