Event Representation-Based Orbit Determination Across Unknown Space Events

摘要

A new way of maintaining the orbit determination across unknown space events using an event representation is presented. Analytic study reveals that the propagated state transition matrix mapping from an initial to a final state is found to be quasi independent of event dynamics on the condition that they drive the initial state to the same final state at a given time. Applying this fundamental finding, numerical simulations verify that a dynamics representation interpolating two disparate' orbit trajectories can be used to linearly propagate the orbit uncertainty information through the state transition matrix. With this property, a batch filter with an event representation is implemented to estimate the orbital state and its uncertainty across unknown space events by applying the thrust Fourier coefficients to represent the event dynamics. A case study with simulated tracking data shows that the batch filter with thrust Fourier coefficients allows one to make use of the previous orbit information to improve the quality of an orbit solution after unknown events. The proposed batch filter has an advantage of avoiding the difficulty of manual tuning of parameters and is able to provide a more accurate postevent orbit determination solution than a regular batch filter solution given limited tracking data.