A PARTICLE FILTERING APPROACH TO SPACE-BASED, MANEUVERING SATELLITE LOCATION AND ESTIMATION

In the highly cluttered environment of frequently used Low Earth Orbit (LEO) and Geosynchronous Equatorial Orbit (GEO), the need for heightened space-based Space Situational Awareness (SSA) capabilities is an increasingly important aspect of maintaining superiority in the space domain. This paper lays out the mathematical framework for a probability-based automated search and estimation algorithm in the case of an actively maneuvering target satellite. Specifically, the case of a space-based observer in a relative reference frame is considered. A particle filter (PF) is used to estimate and update an approximation of the target satellite probability density function (pdf) in the presence of highly nonlinear observation and dynamic models. Reachability set theory for impulsive maneuvers is applied in conjunction with the PF algorithm to alleviate some of the computational burden associated with classical PF’s providing a feasible and accurate option for nonlinear scenarios. Attitude control of the observer spacecraft is formulated as a greedy maximum likelihood optimization problem for computational feasibility. Simulations for both LEO and the GEO cases are presented and discussed in detail.