Date of Award:


Document Type:


Degree Name:

Doctor of Philosophy (PhD)


Mechanical and Aerospace Engineering

Committee Chair(s)

Geordie Richards


Geordie Richards


David Geller


Stephen Whitmore


Douglas Hunsaker


Charles Swenson


The thesis of this dissertation proposes a novel filter algorithm to improve tracking and catalog maintenance of uncooperative satellites and other Resident Space Objects (RSOs) in Geosynchronous Equatorial Orbit (GEO). Tracking can be supported by space-based tracking from observer satellites (OBSs). Practical limitations can lead to long time gaps between measurement updates when tracking RSOs from an OBS, which may induce a loss of fidelity or divergence of the estimation algorithm. The Extended Kalman filter (EKF) is commonly used for tracking RSOs but it diverges as a consequence of nonlinearity in the dynamics and nonlinearity in the optical measurements from OBSs. Both nonlinear- ities cause the underlying probability density (PDF) of the state vector to deviate from a Gaussian distribution, leading to divergence after measurement update. The Unscented Kalman filter (UKF) and the Gaussian Mixture Model filter (GMM) were proposed to solve the divergence problem in the EKF. A hybrid algorithm, the Hybrid Kalman-particle filter (HKF), was developed and likewise assessed to improve on the EKF methodology by com- bining with particle filtering techniques. Lastly, this work presents a novel filter algorithm, the Extended Step-Back Kalman filter (ESBKF), in which the measurement update is ap- plied at a time in the past when the distribution of the RSO in state-space is approximately Gaussian. The filter statistics are then propagated forward to the present, and the nonlinear effects of the dynamics are dramatically reduced, thereby avoiding divergence longer.