This thesis explores the design of pulsed drag-free systems, in terms of experiment control and data analysis. Application cases show that beyond their feasibility, such systems could also provide with valuable scientific results.
Space-borne drag-free systems allow investigating the motion of free-falling test-masses, usually for scientific purposes. The masses are housed inside one or more spacecrafts, as part of the payload. In the concept of this thesis, the relative displacements of the test-masses are controlled by short periodic electrostatic pulses, between which an enhanced free-fall is possible by switching off the actuator. The periodic pulses cause gaps to appear in the science time-series.
First, the control-design problem is formulated for a time-periodic pulsed SISO system as a control-optimum. The setup is then extended to MIMO applications by specifying independent control coordinates for the spacecraft and the test masses.
The data analysis must process time-series with a high spectral dynamic range and periodic gaps. Algorithms based on binned spectrum densities are detailed, for spectral estimation, linear parameter estimation, and yielding unmatched results for gap-filling.
Finally the data analysis is demonstrated on three typical problems. For one the estimation of a force gradient in the LISA Technology Package, matching well the performance of the baseline experiment, which uses a continuous actuation. Then a geodesy mission using a gradiometer near Mars is studied, illustrating how the periodic gaps affect the recovery of a gravity field. Finally, the problem is adapted to LISA, and it is shown which classes of astronomical signals are affected by the periodic gaps.
Adrien Grynagier
data analysis drag-free formation flying gradiometer power spectral density space-borne