Construction of a dynamic adaptive optics system based on the subspace identification

Abstract

Adaptive optics (AO) systems have been used in many applications, such as ground-based astronomical telescopes for improving the resolution by counteracting the effects of atmospheric turbulence. However, the traditional AO system model is not good at dealing with noise interference in the closed-loop correction. Therefore, our paper proposes a subspace system identification method based on errors-in-variables to build an accurate dynamic model of the AO system from measurement data with closed-loop system identification. Experimental simulation results show that the root mean square of the residual wavefront is smaller than that of the traditional method, whether photon noise and camera readout noise exist or not. We conclude that the identified model has good accuracy and noise disturbance compensation ability to deal with dynamic wavefront correction compared with the traditional method.