Design and implementation of a relative state estimator for docking and formation control of modular autonomous spacecraft

Abstract

Modularity is a promising design concept for space systems. In a modular satellite, the individual subsystems would be broken down into physically distinct modules, which would then dynamically recombine into an aggregate vehicle. This could improve the flexibility and reusability of satellites, and could even enable some mission objectives which are not possible at all with monolithic vehicles. However, modularity requires that some additional new elements be included in the design that are not needed with a monolithic satellite. Two of these are a docking interface to allow modules to attach, and a position measurement system to allow modules to fly accurately in formation and dock with each other. These two additional elements are explored in this thesis. The central focus is on a relative state estimator based on an extended Kalman filter. The estimator is first presented theoretically, then the results of implementation and hardware testing are discussed. This thesis presents two main hardware applications for the estimator, both of which mirror prime space-based applications of modularity itself: docking and formation maintenance/reconfiguration.