The unscented Kalman filter applied to satellite orbit determination using only publically available two-line element sets

Abstract

Satellite Orbit Determination (OD) is the process to estimate the position and velocity (the state vector, state, or ephemeris) of an orbiting object. Any satellite in its orbit faces many perturbation forces that affect its position in its orbit and therefore affect the tracking system. This leads to search for a method to correct these drifts. The nonlinear problem of tracking and predicting where a satellite will be over some time is one of OD early challenges. This report shows the design and implementation of an unscented Kalman filter for nonlinear estimation applied to satellite orbit determination, the way it was embedded and the effects of adding it on the tracking results to get more accurate orbital estimation, by correcting the tracking data (from Two-Line Element, TLE, data sets) itself rather than using any external source of measurement for comparison and correction (GPSs, optical sensors, radars...etc.).