The GPU-based High-order adaptive OpticS Testbench

Byron Engler,Markus Kasper,Serban Leveratto,Cedric Taissir Heritier,Paul Bristow,Christophe Verinaud,Miska Le Louarn,Jalo Nousiainen,Tapio Helin,Markus Bonse,Sascha Quanz,Adrian Glauser,Julien Bernard,Damien Gratadour,Richard Clare
2024-11-08
Abstract:The GPU-based High-order adaptive OpticS Testbench (GHOST) at the European Southern Observatory (ESO) is a new 2-stage extreme adaptive optics (XAO) testbench at ESO. The GHOST is designed to investigate and evaluate new control methods (machine learning, predictive control) for XAO which will be required for instruments such as the Planetary Camera and Spectrograph of ESOs Extremely Large Telescope. The first stage corrections are performed in simulation, with the residual wavefront error at each iteration saved. The residual wavefront errors from the first stage are then injected into the GHOST using a spatial light modulator. The second stage correction is made with a Boston Michromachines Corporation 492 actuator deformable mirror and a pyramid wavefront sensor. The flexibility of the bench also opens it up to other applications, one such application is investigating the flip-flop modulation method for the pyramid wavefront sensor.
Instrumentation and Methods for Astrophysics
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to develop and evaluate new control methods for extreme adaptive optics (XAO), especially for the Planet Camera and Spectrometer (PCS) on the European Southern Observatory (ESO)'s Extremely Large Telescope (ELT). Specifically, the paper introduces the GPU - Accelerated High - Order Adaptive Optics Test - bed (GHOST), which aims to study and evaluate new control methods, such as machine learning and predictive control. These methods are crucial for the XAO systems of future large telescopes like the ELT. ### Main problems: 1. **Requirements for extreme adaptive optics (XAO)**: - The PCS of the ELT requires extremely high - contrast imaging capabilities to directly image and characterize nearby exoplanets. For this purpose, the XAO system must be able to effectively correct wave - front errors caused by atmospheric turbulence and needs to handle a large number of modes (for example, about 4,000 modes in the first stage and about 10,000 modes in the second stage). - Traditional integral controllers have limitations in handling high - speed, high - precision wave - front corrections. Therefore, more advanced control methods, such as predictive control and machine - learning - based control, need to be developed. 2. **Advantages of predictive control**: - Predictive control uses past measurement data to predict the wave - front state at the current moment, thereby reducing time - delay errors and possibly reducing the influence of photon noise. - Compared with traditional controllers, predictive controllers can better cope with rapidly changing wave - front errors and improve the performance of the system. 3. **Design and implementation of GHOST**: - GHOST is a two - stage XAO system. The first - stage correction is carried out in a simulated environment, using a pyramid or Shack - Hartmann Wave - Front Sensor (Shack - Hartmann WFS), controlling 800 - 1,000 modes at a frequency of 1 kHz. - The second - stage correction uses a 492 - actuator deformable mirror from Boston Micromachines Corporation and a pyramid wave - front sensor, controlling 300 - 400 modes at a frequency of 2 - 4 kHz. - The design of GHOST allows researchers to inject simulated first - stage residual wave - front errors and perform second - stage corrections on actual hardware, thereby verifying the effectiveness of new control methods. 4. **Other applications**: - GHOST also has flexibility and can be used for other experiments, such as studying the flip - flop modulation method and exploring how to quickly start and stop the modulator in the pyramid wave - front sensor. ### Summary: This paper solves the problem of how to develop and evaluate new control methods suitable for XAO systems of future large telescopes such as the ELT by introducing the GHOST platform. GHOST not only provides an experimental platform for predictive control and machine learning but also provides a tool for research in other related fields.