Surrogate model-based wavefront sensorless adaptive optics system for correcting atmospheric distorted images

Atmospheric turbulence is the major reason for blurring of the images from ground-based telescopes. In this paper, a model-based wavefront sensorless approach to remove the distortion from the images is proposed. A radial surrogate model is used to simplify the optimisation of an intensity-based objective function. To reduce the number of intensity evaluations required to build the model, the statistics of the atmosphere are extracted and explored. The number of evaluations can be further decreased by employing space-filling sampling and sub-sampling plans. The accuracy of the proposed approach is improved by optimising the associated hyper-parameters. This method is shown to efficiently remove the first 8 aberration modes after tip and tilt within the maximum number of evaluations allowed. Results indicate that applying the proposed method to a random atmosphere phase screen can improve the fractional encircled energy by about 20%.