Maneuver Detection with Event Representation Using Thrust Fourier Coefficients

A systematic way of detecting unknown maneuvers is developed by representing an unknown acceleration tied to an event with thrust Fourier coefficients. Event representation using thrust Fourier coefficients can rigorously represent an unknown maneuver by generating an equivalent maneuver with the same secular behavior. By appending 14 thrust Fourier coefficients as solve-for states, the modified sequential filter processes observation data both forward and backward in time to detect maneuver onset and termination time, respectively. Along with the represented perturbing acceleration, the detection algorithm provides more accurate postmaneuver orbit solutions. A case study of detecting unknown maneuvers with different types of simulated measurement data verifies the presented approach.