Research on telescope alignment using a generic wavefront sensorless method based on images with different sampling parameters

Abstract

Telescope alignment is one of the key factors to ensure the image quality and resolutions. In this paper, a wavefront sensorless method based on images using optimization algorithms is proposed. To verify the validity and universality, different kinds of telescopes are employed, and misalignments are corrected based on images with various sampling parameters. Nesterov-accelerated adaptive moment stochastic parallel gradient descent (Nadam SPGD) algorithm is adopted to search for the optimal solution in simulations. Results show that the average errors in the 80% encircled energy radius between corrected images and designed images are no larger than 0.2 pixels, satisfying the demands of image quality. Without increasing complexity of optical systems and alignment processes, the proposed method can align telescopes effectively even if images are undersampled seriously.