SPRINT, system parameters recurrent invasive tracking: a fast and least-cost online calibration strategy for adaptive optics

Future large adaptive telescopes will trigger new constraints for the calibration of adaptive optics (AO) systems equipped with pre-focal deformable mirrors (DMs). The image of the DM actuator grid, as seen by the wavefront sensor (WFS), may evolve during the operations because of the flexures of the opto-mechanical components present in the optical path. This will result in a degraded AO performance that will affect the scientific operation. To overcome this challenge, it will be necessary to regularly monitor and compensate for these DM/WFS mis-registrations, either by physically realigning some optical components or by updating the control matrix of the system. In this paper, we present a new strategy to track mis-registrations using a pseudo-synthetic model of the AO system. The method is based on an invasive approach where signals are acquired on-sky, before or during the scientific operations, and fed to the model to extract the mis-registration parameters. We introduce a method to compute the most sensitive modes to these mis-registrations that allows us to reduce the number of degrees of freedom required by the algorithm and to minimize the effect on the scientific performance. We demonstrate that, by using only a few of these well-selected signals, the method provides very good accuracy for parameter estimation, well under the targeted accuracy, and has a negligible effect on the scientific path. In addition, the method appears to be very robust to varying operating conditions of noise and atmospheric turbulence and it performs equally for both pyramid and Shack-Hartmann WFSs.