Multi-layer predictive control for tomographic wavefront estimation

We discuss the extension of predictive control techniques utilizing Taylor frozen-flow motion to multi-guide star, multi-layer tomographic wavefront measurement systems. The expectation is that the combination of height information from multiple guide stars and wind velocity estimates breaks degeneracies in volumetric phase reconstruction, particularly for voxels sampled by only one GS, allowing for a reduction of the tomographic error and an expansion of the field of view. Using a simple shifting-and-averaging scheme to track individual layer motions in simulation, we demonstrate 3-10% reductions in the tomographic wavefront estimation error of individual layers for an aperture size of 10 meters, subapertures of 30 cm, a Mauna-Kea type atmospheric profile, and wind velocities of 10 m/s. The majority of the benefits occur in regions sampled by only 1-2 LGS’s downwind at high altitudes. An idealized scenario with 100% Taylor frozen-flow motion, perfect knowledge of wind velocities, and noise-free wavefront sensors is assumed.