Reconstruction Strategies for GeMS

Abstract

Abstract. The Gemini Multi-conjugate adaptive optics System (GeMS) is a facility instrument for the GeminiSouth telescope that can provide near-uniform atmospheric compensation over a 1 arc minute square field of view. To provide diffraction-limited image quality at near IR wavelengths across an extended field-of-view, the system includes 5 laser guide stars, 3 natural guide stars and 3 deformable mirrors optically conjugated at 0, 4.5 and 9km. By many ways, GeMS represents a unique system to prepare and understand any future wide field AO systems planned for the ELTs. In this paper, we present the different strategies developed for the phase volume estimation and correction relative to the LGS high-order path. We compare the performance obtained by a Least-Square Estimator (LSE) and a Minimum Mean Square Error (MMSE) reconstructor. The MMSE reconstructor is found to improve the performance compared to the classical LSE, particularly when the system is affected by noise. We show the results obtained by realistic simulations as well as the first results obtained in laboratory on CANOPUS, the GeMS AO bench.

A Sample of GeMS Calibrations and Control Schemes

“…Recently, we have interfaced a CCD camera in parallel to the DWFS presented above. This will allow us to start investigating the dynamical performance of the benchand rank our various reconstructors [4],-validate our synthetic iMat approach, and extend/check the NCPA analysis. The latest schedule puts the first light in June 2010.…”

“…It has been included for reference and to illustrate the complexity of this system. A more complete description of the system can be found in [4,5]. ↓ It can be argued that to fully include real system effects, the iMat would need to be recorded on the sky, that is, in the actual conditions it will be used, and a lot of work has been done in that direction [6].…”

On-sky tests of the CuReD and HWR fast wavefront reconstruction algorithms with CANARY

“…It is known that LSE is not the optimal control algorithm and that performance can be improved by using regularization, for example with the minimum mean square error or with the maximum a-posteriori approach, see for example (Flicker, Rigaut & Ellerbroek 2000;Le Louarn 2002;van Dam, Le Mignant & Macintosh 2004;Petit et al 2005;Neichel et al 2010). With such a procedure, additional knowledge is built into the control matrix about the expected shape that the DM needs to correct.…”

The Gemini MCAO System GeMS: nearing the end of a lab-story

“…In Fig.11 right-side, we come back to the original 4layers Cn2 profile, and once this atmospheric change has been detected, the first command matrix is reloaded. We have implemented different command matrix types (Least-Square, Weighted Least-Square, Minimum Mean Square Error, see Neichel et al 23 ) so the optimizer would also have to choose the most appropriate one. …”