Sparse sampling algorithm for deformable mirror control

Abstract

For the efficient and accurate control of deformable mirror (DM) shapes, we propose a method of sparse sampling of control parameters to attain accurate control of the high- and low-order coefficients of the DM surface shape expressions. The influence function is used and modified to reduce the hysteresis effect and provide an accurate solution using the statistical multiple regression method. A test system was built and performed the contrast fit experiments for the different DM surface shapes using the proposed algorithm and the conventional method. The proposed algorithm can significantly improve the convergence speed of the DM surface shape, achieving the maximum convergence residual of 1/20λ.