Control system design for spacecraft formation flying: Theory and experiment

Spacecraft formation flying is an enabling technology for many future space science missions, such as separated spacecraft interferometers (SSI). However the sensing, control and coordination of such instruments pose many new design challenges. SSI missions will require precise relative sensing and control, fuel-efficient, fuel-balanced operation to maximize mission life and group-level autonomy to reduce operations costs. Enabling these new formation flying capabilities requires precise relative sensing and estimation, enhanced control capabilities such as cooperative control (multiple independent spacecraft acting together), group-level formation management and informed design of a system architecture to manage distributed sensing and control-system resources.; This research defines an end-to-end control system, including the key elements unique to the formation flying problem: cooperative control, relative sensing, coordination, and the control-system architecture. A new control-system design optimizes performance under typical spacecraft constraints (e.g., on-off actuators, finite fuel, limited computation power, limited contact with ground control, etc.). Standard control techniques have been extended, and new ones synthesized to meet these goals.; In designing this control system, several contributions have been made to the field of spacecraft formation flying control including: an analytic two-vehicle fuel-time-optimal cooperative control algorithm, a fast numeric multi-vehicle, optimal cooperative control algorithm that can be used as a feedforward or a feedback controller, a fleet-level coordinator for autonomous fuel balancing, validation of GPS-based relative sensing for formation flying, and trade studies of the relative control and relative-estimation-architecture design problems. These research contributions are mapped to possible applications for three spacecraft formation flying missions currently in development. The lessons learned from this research have been validated in simulation and experiment.