MEMS adaptive optics for the Gemini Planet Imager: control methods and validation

The Gemini Planet Imager (GPI) Adaptive Optics system will use a high-order MEMS deformable mirror for phase compensation. The MEMS mirror will be used in a Woofer-Tweeter configuration, with a frequency-domain based splitting of the phase between the two mirrors. Precise wavefront control depends on the ability to command them MEMS to make the exact phase desired. Non-linearities in the MEMS may prevent this. We determine that influence-function pre-compensation can remove most, but not all, open-loop error. We use simulation and a simulation of a non-linear MEMS to address the issue of how much non-linearity can be tolerated in closed-loop by GPI.