Parallel Estimation and Control Architectures for Deep-Space Formation Flying Spacecraft

Precision formation flying in deep space can be used to implement optical instruments capable of imaging planets in other solar systems. The distance of the formation from Earth necessitates a significant level of autonomy. Each spacecraft must base its actions on its estimates of the relative position and velocity of the other spacecraft. To achieve the required precision coordination and control, the flow of information between spacecraft must be carefully designed and novel techniques developed for the design of optimal onboard estimators. The use of standard Kalman filter-based designs can lead to unanticipated dynamics-which we refer to as disagreement dynamics-in the estimators´ errors. We show how communication amongst the spacecraft can be designed in order to control all of the dynamics within the formation. We present several results relating the topology of the communication network to the resulting closed-loop control dynamics of the formation.