自适应光学
波前
计算机科学
梯度下降
次镜
光学
波前传感器
主镜像
变形镜
望远镜
物理
人工智能
人工神经网络
作者
Min Li,Xin Liu,Ang Zhang,Junbo Zhang,Hao Xian
标识
DOI:10.1109/lpt.2024.3363210
摘要
The precise secondary mirror alignment is a key factor to ensure the imaging quality of a segmented telescope with a high resolution demand. In this paper, a two-step alignment method using an improved stochastic parallel gradient descent (SPGD) algorithm called Nadam SPGD algorithm is proposed. Simulation results show that our method can distinguish aberrations caused by phase errors of the primary mirror and position errors of the secondary mirror precisely, and align the secondary mirror quickly and accurately. The improved SPGD algorithm can enhance the correction efficiency by more than two times compared with the traditional SPGD algorithm.
科研通智能强力驱动
Strongly Powered by AbleSci AI