Nonlinear Filtering Using Directional Statistics for the Orbital Tracking Problem with Perturbation Effects
John T. Kent, Shambo Bhattacharjee, Islam I. Hussein, Moriba Kemessia Jah · 2018
In this paper we consider the space object tracking problem to predict the state of an orbiting object from a sequence of angles-only measurements. Using ideas from directional statistics, an “Adapted STructural (AST)” coordinate system has been developed to represent the state vector, and in this coordinate system, the solution to the tracking problem effectively reduces to the standard Kalman filter. In this paper we investigate the filter in more detail, both to consider a wider variety of orbital regimes and to incorporate perturbation effects