31 December 2008 Adaptive optics correction experiment based on stochastic parallel gradient descent algorithm
Author Affiliations +
Proceedings Volume 7130, Fourth International Symposium on Precision Mechanical Measurements; 71303Y (2008); doi: 10.1117/12.819701
Event: Fourth International Symposium on Precision Mechanical Measurements, 2008, Anhui, China
Abstract
The stochastic parallel gradient descent (SPGD) algorithm is a promising control algorithm for adaptive optics (AO), which is independent of wave-front sensor and is used to correct the wavefront distortion by optimizing the system performance metric directly. In this paper an adaptive optics experiment system with 32 control channels is set up and the static phase-distortion correction experiment is carried out to research the efficiency and performance of SPGD algorithm. The quadratic sum of intensity and the mean radius are used as the system performance metric respectively in the experiment. The experiments results demonstrate that the adaptive optics using SPGD algorithm can correct the static phase-distortion successfully. The mean radius is more sensitive to the small perturbation voltage than quadratic sum of intensity. When the mean radius is used as the system performance metric at the beginning of the correction process, and then the encircled energy is used in succession, both the convergence rate and the stability are improved.
© (2008) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Bo Chen, Huizhen Yang, Xinyang Li, Wenhan Jiang, "Adaptive optics correction experiment based on stochastic parallel gradient descent algorithm", Proc. SPIE 7130, Fourth International Symposium on Precision Mechanical Measurements, 71303Y (31 December 2008); doi: 10.1117/12.819701; https://doi.org/10.1117/12.819701
PROCEEDINGS
6 PAGES


SHARE
KEYWORDS
Adaptive optics

Control systems

Stochastic processes

Wavefronts

Optical calibration

Computing systems

Distortion

Back to Top