Interacting Multiple Model Estimation for Spacecraft Maneuver Detection and Characterization
Sangjin Lee, Inseok Hwang · 2015
As the number of spacecrafts performing orbital maneuvers has been increasing, the timely detection of the maneuver along with post-maneuver orbit characterization has become crucial for space situational awareness. In this paper, a new algorithm to detect and characterize maneuvering spacecrafts is proposed based on the interacting multiple model estimation. It is assumed that the types of possible maneuvers are finite and each maneuver is characterized by resulting change in the orbital elements. Under the assumption, the maneuvering spacecraft is modeled as a jump Markov nonlinear system having multiple modes, each of which describes the spacecraft’s motion subject to the corresponding orbital elements. Through the interacting multiple model estimation, the mode probabilities which represents how likely each mode is currently in operation are computed. The maneuver detection and the post-maneuver orbit characterization are then performed simultaneously by identifying the mode with the highest mode probability. The effectiveness of the proposed algorithm is demonstrated through numerical simulations.