Optimization Of Adaptive Optical System Based On Imaging Information

Light are affected by internal and external factors during transmission.Both of these factors will cause wavefront distortion of light to produce wavefront aberrations,and reduce the quality of reception and transmission efficiency.At present,Adaptive Optics(AO)technology is an effective method to solve wavefront distortion.The AO system based on imaging information has the advantages of simple structure and wide application fields,and has become a research hotspot in the related fields of adaptive optics in recent years.Stochastic Parallel Gradient Descent(SPGD)has the advantages of parallelization,strong robustness,and simple implementation.It is the most promising one of AO system control algorithms based on imaging information.Optimal control algorithm is a common way to optimize AO system based on imaging information.This thesis studies and analyzes the inherent limitations of the conventional SPGD algorithm and proposes corresponding optimization methods for verification.(1)Regarding the conventional SPGD algorithm,the driving voltage of the DM is directly used as the control parameter,which is likely to cause the contradiction between the algorithm’s convergence speed and the correction ability of the DM,so a SPGD algorithm based on mode is proposed.The optimization algorithm uses Zernike mode coefficients that can change with aberrations as control parameters,making the algorithm more flexible and improving the convergence speed of the algorithm.The simulation results show that the system convergence speed of the SPGD algorithm based on mode is increased by more than 50%.Then build an AO system experimental platform based on imaging information,and use the optimized SPGD algorithm to correct the first 10 and the first 35 aberrations in the experiment.The experimental results show that the convergence speed of the optimized system is increased by about 50%,and the calibrated spot reaches the diffraction limit resolution,indicating that the system has a good correction capability.(2)In the conventional SPGD algorithm,the gain coefficient is a certain fixed value.In special cases,it is easy to reduce the convergence speed of the algorithm and increase the probability of falling into a local extreme value.Combining the Adam(Adaptive Moment Estimation)algorithm,which is commonly used in deep learning,has the characteristics of adaptive gain adjustment,the SPGD algorithm of adaptive gain is proposed.The comparison and analysis of AO system based on SPGD algorithm based on adaptive gain and conventional SPGD algorithm are carried out.The simulation results show that the system convergence speed based on the adaptive gain SPGD algorithm is increased by about 50%,and the probability of falling into a local extreme value is reduced by about 30%.In summary,this thesis proposes corresponding optimization methods for the shortcomings of the traditional SPGD algorithm and performs simulation verification,and then builds an experimental platform to verify the feasibility and effectiveness of the SPGD algorithm based on mode.The experimental results in this thesis provide a theoretical basis and application reference for AO system optimization based on imaging information.